B747 strange FMC climb gradient change

[Guest lvedin]

PMDG Boeing 747-400.Due to I started laborate to understand how to count and effect climb gradient up to a level, using FMC as help to include relevant parameters. After a lot of trying settings, I found for me total not explainable behavours.IN SUMMARY:PAGE TAKEOFF REF L1: FLAPS 10/905FTFMC indicate speed/alt 176/9657 after 20nmPAGE TAKEOFF REF L1: FLAPS 10/900FTFMC indicate speed/alt 327/13887 after 20nmHOW STRANGE ??? MAJOR change in climb gradient just by change flaps retract by 5 FT.Repeatable by change PAGE TAKEOFF REF L1: 10/905FT > 10/900FT > 10/905FT repeated. Repeatable by restart FS.Can anyone explane this behavour, why is 900FT magic and how can possible the performance change such drastic ?Following is detailed FMC setting after cold startup:POS:ESSAROUTE:ESSA KAS EDDMPERF:Payload 122Grossweight 680ZFW 516 RESERV 24COST 80CRZ ALT FL300THRUST LIMIT: nonePAGE TAKEOFF REF:L1: FLAPS 10/1000FTDEPARTURE:runway 19RPAGE TAKEOFF REF:execute all speed reference buttons R1-4now enter a fix to examine.new FIX:ESSA BRG/DIS 190/20 download the fix to scratch-board and insert as the first fix in the page LEGS as "ESSA01".(i.e. the first fix in LEGS is ESSA01 20nm in rwy-heading)PAGE LEGS indicate:ESSA01 <CTR> 176/9728 (in addition the speed 176 is strange since it indicate no acceleration yet)now go to PAGE TAKEOFF REFL1: FLAPS 10/905FTPAGE LEGS indicate no change, OK:ESSA01 <CTR> 176/9728 now go to PAGE TAKEOFF REFL1: FLAPS 10/900FTPAGE LEGS indicate unrealistic performance change:ESSA01 <CTR> 327/13887/ Lennart Vedin

[downscc 1,327]

You are taking two data points on two different performance profiles and drawing conclusions without evaluating to whole performance profile.Without running trials, my guess is that your second data point at 20 nm has more variance than expected because in one case it is below 10,000 and the other it is above 10,000. If you can visualize what that time/distance/altitude profile looks like then you notice there is a discontinuity where the speed restriction is removed. I think the variance would make more sense if you also look at the elasped time from Vr to ESSA01 and added a few more data points at distance.I ran many weight vs takeoff trials last year, which took several weeks, and found that the time to climb to 10,000 is most sensitive to GTOW. I don't think the acceleration height is going to make an unexpected difference. I think you just happen to pick a point that was below 10,000 in one case and above in the other. I think if you also looked at the climb to FL180 the data would make more sense.

[Guest lvedin]

I only changed TAKEOFF REF L1: FLAPS 10/905FT to TAKEOFF REF L1: FLAPS 10/900FT in the two huge different cases. I do not change anything else.Do I understand this setting wrong TAKEOFF REF L1 is at what alltidude the flaps is started to be retracted by crew.Default TAKEOFF REF L1 is xx/1000FT. And even is this case 10/1000FT why is still ESSA01 <CTR> 176/9728 as like the flaps is not expected to be retracted at all.It must bee some major fault how I handle the FMC, while the FMC behavour seams quite strange since only what should be happen at 900FT or 905FT, has nothing to do with speed restriction, speed restriction is the same and the performance difference is huge, probalbly due to the first case do not take no flaps retraction at 905FT (or default 1000FT).Else I have successfully modify TAKEOFF REF L1.

[Guest lvedin]

A quick acceleration above FL100 I do understand, and does not indicate major performance difference as you correct explain.I should have add aslo, it is NOT the speed difference that is remakable (i just notice that the second higher profile at least not have lower speed at ESSA01), it is as the title say the climb gradient difference that is remakable.With TAKEOFF REF L1: FLAPS 10/1000FT, a typical clean climb-speed should be around 250kts, not 176kts all the way to 9657ft.

[downscc 1,327]

Your clean climb below 10,000 should be 250 or higher... don't know what you are doing, maybe a screen shot.Just as a side note, no airline is going to lower the nose for acceleration below 1000 agl, I bet most keep the high angle of climb higher than that. If you don't have runway ahead you want all the altitude below you can get.

[Guest lvedin]

Thanks for answer.I did find one source explaining why the settingTAKEOFF REF L1: FLAPS 10/905FT to TAKEOFF REF L1: FLAPS 10/900FT make so much differnt.FLAPS 10/901FT up to 10/1000FT is same counting. I.e. the value is somehow rounded up to 1000FT. While 10/900FT is not rounded.I did post a new topic with picture, since ther is an other seams more clear why to show the strange behaviour.http://forums1.avsim.net/index.php?showtopic=243391