FS2004 [ Solved ] 747 Trim Numbers

[FSMuseum]

Hi all,

Continuing work on my 747, I have stumbled into a problem that I struggle to find a solution for.

I want to setup the pitch trim to perform accurately according to the readings on the RFP 747-200 panel. This panel's trim gauge defaults to 7.5 units no matter what the AIR file has in it, and both the panel and the real 747 have a trim range of 15 units.

The green band for takeoff ranges 2.5 to 9.5 units, and based on some rudamentary guessing, math, and CG data, this correlates to the following takeoff trim settings based on the takeoff CG limits:

CG 33% MAC = 2.5 Units
CG 13% MAC = 9.5 Units

With some interpolation, I can determine that 7.5 units should correlate to ~18.75% CG.

Using this information, I made a saved flight with my 747 in standard conditions with a CG of 18.75% and the trim left at the starting position of 7.5. By trial and error, I can set the Cm0 Trim 0's value such that taking off in this configuration results in a nice easy climb with only very gentle back pressure to keep the nose up that can be trimmed out with just a click of the trim switch or two.

However, I can't seem to set the effectiveness, or hardly any other values or tables, correctly according to real world numbers.

Using a Cm_dt value of -2.5, and Cm0 of -0.05, causes some undesirable behavior at the extreme CG ends.

33% CG with correct trim results in a very heavy nose down moment. The elevator can easily overpower it, but letting go will cause the aircraft to pitch down more than about 3 - 4 degrees per second continuously, which seems like too much nose down moment.

13% CG with correct trim is not terrible, but more than a few pounds of back pressure causes the aircraft to pitch up excessively, which also doesn't seem realistic.

Further, in cruise with a CG of 31%, the real aircraft will trim to about 2.5 - 2.7 units. My 747, however, trims to about 5.7 - 6.2, which of course is not realistic. I assume that once I fix the takeoff trim, I can use the Delta_Cm0 due to mach graph to adjust the cruise behavior(?)

Studying Yves's document doesn't seem to yield much in the way of help for me. His equation for Cm_dt doesn't entirely work for FS9 because FS9 does not use table 1541, and beyond that his term "delta_etr [elevator_trim_limit] is confusing to me. Does this mean substitute the full travel of the stabilizer (15 units)? Or does this mean to substitute a particular value for the trim (such as the difference between a particular trim setting and 0 trim, such as 9.5 - 7.5 = 2.0)?

 

[Roy Holmes]

Starting at the end, 1541 post dates FS9 but the data therein is contained in 1101
In the sim the trim system is mechanised in a similar way as in the 747, a moving stabilizer and elevators. The elevator trim limit is what is in the aircraft.cfg.
In the sim all angles are actual radians or degrees and NOT units.
Units are typically used to give correlation with a gauge indications and where the units on the gauge are not exactly degrees or even linear when measured at the stabilizer.
This means that you can not strictly compare trim degrees to stab units. This could be one of your issues.
Your trim and stability are affected by static margin. this should be in the region of 10% for a transport airplane. It is the ratio of the Cm slope divided by the CL slope. The Cm slope you will find in 473, CL in 404. Typically you may have to adjust 473 to get a good Static Margin.
Lastly the aerodynamics coefficients for 747 are given in Appendix C6 Airplane F of J Roskam "Airplane Flight Dynamics and Automatic Flight Control Part 1.
Before using those, check Yves document for the differences in sign convention between real world and sim. Your signs above are correct.
Roy

 

[FSMuseum]

Hi Roy,

Thanks for your input!

I am aware of the difference between units and degrees. With Boeing, the conversation is simple on the 747 classic series. 0 Units is -2 degrees of actual stabilizer angle, and 15 units is +13 degrees.

Having looked at all that I can find, including real Boeing test data for Cm and CL, the real static margin of the 747 is around 20 - 30% actually (the previous documents I was looking at were blatantly wrong and somehow I missed that) this being due to the enormous size (and therefore enormous CG range) of the aircraft. The neutral point exists at about 48% MAC with a typical empty CG of 25%. My 747, with accurate Cm and CL graphs, indicates about 24.3% in AirWrench, and the actual pitch behavior is really nice.

My problem comes in figuring out how to calculate the proper values for how much pitch moment the stabilizer should give. Roskams documents don't cover this as far as I can see/find, and Yves's document still confused me here.

 

[Roy Holmes]

My problem comes in figuring out how to calculate the proper values for how much pitch moment the stabilizer should give. Roskams documents don't cover this as far as I can see/find, and Yves's document still confused me here

My take on this is that you should provide sufficient trim to cover the flight envelope at all combinations of weight and CG. That would be what Certification pilots look for.
Or another method would be to fly in autopilot with Alt hold selected and you should not find it running out of trim, especially at slow speeds near stall with flap when the UP limits will be tested. At the stall with flap the 747 has a straight nose drop. With flaps up it does not and the stall is defined by the buffet level felt in the cockpit. That could be hard to simulate

Yves Stabilizer pitch moment expression is CMdt (in 1101) times trim range (in .cfg) times 536 times 1525 times the elevator trim effectiveness scalar in .cfg.
473 usually gives zero at zero AOA which is strictly incorrect, however CM0 is added and that then gives the necessary nose down moment at zero AOA.
Roy

 

[FSMuseum]

My take on this is that you should provide sufficient trim to cover the flight envelope at all combinations of weight and CG. That would be what Certification pilots look for.
Or another method would be to fly in autopilot with Alt hold selected and you should not find it running out of trim, especially at slow speeds near stall with flap when the UP limits will be tested. At the stall with flap the 747 has a straight nose drop. With flaps up it does not and the stall is defined by the buffet level felt in the cockpit. That could be hard to simulate

Yves Stabilizer pitch moment expression is CMdt (in 1101) times trim range (in .cfg) times 536 times 1525 times the elevator trim effectiveness scalar in .cfg.
473 usually gives zero at zero AOA which is strictly incorrect, however CM0 is added and that then gives the necessary nose down moment at zero AOA.
Roy

Since my problem is finding what value to use for Cmdt, could I set up an equation like this: find total pitch moment at a particular configuration without stabilizer influence, then set up a new total pitch moment equation with Cmdt added as a varibable, solve for Cmdt, to get the proper Cmdt?

 

[FSMuseum]

Hi again,

Just to report back, I solved my problem. Turns out I had made a typo when entering my Cm graph into text using AirUpdate, resulting in a graph that was just slightly not steep enough matched against the real Boeing graph.

I changed the graph, adjusted the flap pitch moments, and then lowered the Cmdt to a value I calculated using the above equation I mentioned, slightly adjusting Cm0 to account for the RFP trim starting at 7.5 units, and wallah!

Takeoff behavior is now almost exactly as I would expect at both ends of the CG spectrum, and cruise trim is right on the money too! Amazing how some simple errors and redone math can result in a much nicer result...

This 747 is coming closer and closer to completion

Thanks for the help!