|
Post by acourt on Mar 25, 2013 12:06:23 GMT
Guys, I'm curious about slats in FS9, as I know almost nothing (well...nothing, really) about flight dynamics modeling. The DC-9, like other airplanes, generally pitches nose up and requires nose down trim as you extend the flaps. But extending the slats has the opposite effect, requiring nose up trim as you extend them. Is it possible to replicate that in FS9? As I stated in the title, this isn't a complaint. The Nines are fine the way they are. But getting the slats to "feel" right would be a nice bonus
|
|
|
Post by Mike Monce - HJG on Mar 25, 2013 14:42:41 GMT
Well, first of all, I don't think I agree with your basic premise. You are implying that the slats actually produce some negative lift. If you take a look at this document you will see that slats have very little effect on lift (even includes a graph for a DC9!) adg.stanford.edu/aa241/highlift/highliftintro.htmlThat being said, yes, you can adjust the slat lift in FS9 in the flaps section of the cfg. The slats are labeled as type "2" and you can adjust the lift scalar which is a multiplier to a value that exists in the air file. Generally the lift scalar is a relatively small value compared to that used for the trailing edge flaps. Just be aware that in FS, adjusting one parameter does not necessarily mean you will get the desired result as all the aerodynamics parameters are interdependent. You could get the your desired result of trim up with slat extension, but then your T/O and approach pitch attitudes will be buggered (I am picking up some New Zealand phraseology.... Mark is a bad influence ;D ) Mike
|
|
|
Post by christrott on Mar 25, 2013 17:08:16 GMT
Mike, if you look at figure 3 and ignore the flaps then draw a line from the center of the extended leading edge from the slats in the takeoff configuration through where the end of the flap would be in the retracted position (just ahead of the end of the lubber line), you'll see that the wing now has a negative incidence. This causes the nose down tendency.
Now, when you add the flaps to the takeoff position and draw the same line from the leading edge of the slat to the trailing edge of the flap, the new "wing" has a neutral to slightly positive incidence, causing a nose-up pitch. Then, when you move to the landing position, you will see that line has a significant positive incidence. The amount of nose-up tendency this would cause is offset by the fact the wing is now moving much slower meaning both it is creating less lift, requiring less offset.
|
|
|
Post by christrott on Mar 25, 2013 17:14:54 GMT
Oh, and as for FS9 - what you need to adjust is the pitch scalar, not the lift scalar.
In the default DC-9-30 CFG, the slats have this (the section [Flaps.1])-
pitch_scalar=3.040
Try changing this to a value like -1.000, then add the pitch scalar of 3.040 to the flaps section ([Flaps.2]) as it currently does not have a pitch scalar attached. Tweak from there until it "feels" right. This should not change any other performance of the airplane with the way FS9 does its thing, the pitch effect is a separate table and doesn't interrelate (for some stupid reason).
|
|
|
Post by acourt on Mar 25, 2013 18:52:04 GMT
Mike and Chris,
Thanks for the responses! I can assure you that the slats do in fact have the opposite effect on the aircraft's trim as the flaps. I'll give your hints a try later tonight and report back.
Edited because I wrote the wrong name. Sorry Mike!
|
|
|
Post by Mike Monce - HJG on Mar 25, 2013 20:58:09 GMT
Chris,
I think you are reading way too much into Figure 3; it just shows the relative positions of the slats and flaps. I don't think it says anything about angle of incidence... it's just a line drawing for reference of relative positions in different phases of flight. Look at Fig. 9...that's the one that tells the story. As far as I understand, slats do only one thing: increase the range of AoA with various airspeeds by keeping the boundary layer from separating. Now, leading edge flaps are a different story.
Yes, I guess you could change the pitch scalar in FS, but once again I am not convinced that reflects real world behavior.
Mike
|
|
|
Post by acourt on Mar 26, 2013 0:02:36 GMT
Mike, I currently fly 717s for a living. I can assure you that behavior is accurate Extending the slats moves the center of pressure forward, which causes the trim change. Extending the flaps causes the center of pressure to move aft, hence the difference in trim response. I'll try some changes later and report back. Gotta' get the kids in bed first!
|
|
|
Post by acourt on Mar 26, 2013 0:57:47 GMT
Mike and Chris,
I took the information you guys gave me and set to work for a few minutes. Right now, I've changed the slats' pitch_scalar to -10.0, and we're getting close to the real world. The trim change is rather substantial: when you retract the slats on the real airplane, it takes about four seconds of nose down trim to keep the airplane in trim.
I also changed the extension time for the slats to 1.5 instead of 3.0 (the Nine's slats tend to "whang" out there with very little grace). It would be perfect if I could find a way to insert a delay of approximately one second before the slats extended.
Thanks again for the help, Gents. Next time, I'll modify this post with the results. When I'm done, would you guys like to try it?
***
I've been fiddling some more, and I must say I'm pleased. First off, HJG did a marvelous job of replicating the DC-9's behavior. Honestly, I can't believe how close it is in many respects! When loaded properly, it even lands right.
A great way to tell is by using the pitch and power settings from the Flight With Unreliable Airspeed abnormal procedure. The HJG performance is amazingly close to the charts (within 2-3% N1 close in most cases). Well done!
Second, I think I'm close on the trim change with slats. I think I'm going to change the extend/retract time back to 3.0. The change to 1.5, while a little more realistic, just isn't right without the delay I mentioned earlier.
More later!
|
|
|
Post by Mike Monce - HJG on Mar 26, 2013 12:19:38 GMT
Acourt, That sounds encouraging. Also, what is the deployment sequence with respect to slat and flaps in the real plane? i.e. what are the slat angle, flap angle vs flap lever? Is there a table available that shows the relation? Also make sure once you are satisfied with the change in the slats that you check the plane's behavior in all the other regimes and speeds. Just speaking from experience of the many hours Mark and I have spent where a small change one place leads to undesired results in other areas. Flight modeling with MSFS is definitely a game of compromises and sometimes unfortunate tradeoffs. It really is amazing it does as well as it does using parameter look-ups instead of direct calculation like in x-plane. Bruce Artwick and company wrote some really robust code back in the early 1980's Mike
|
|
|
Post by acourt on Mar 26, 2013 16:01:24 GMT
Also make sure once you are satisfied with the change in the slats that you check the plane's behavior in all the other regimes and speeds. Just speaking from experience of the many hours Mark and I have spent where a small change one place leads to undesired results in other areas. Flight modeling with MSFS is definitely a game of compromises and sometimes unfortunate tradeoffs. Man, you aren't kidding! I finally settled on a pitch scalar for the slats of -20.0 (yep, it's that big a trim change). Unfortunately, now I find myself running out of nose up pitch trim with approach flaps (15 degrees) at about 180 knots. I guess I'll back it off a little bit. Here's the quick rundown on the slat/flap relationship... The slats are either extended or retracted. The flaps have detents at 5, 15, 25, 40, and 50 degrees. (The flap positions are different on some models. The 717 is 13, 18, 25, and 40, although a "Flap Takeoff Selector" gives you the ability to set any flap setting you want between 5 and 20 degrees; the MD-80 is similar). Both the slats and flaps are controlled with one handle. The first detent is EXT/0, as in slats extended, flaps zero. From there the detents are EXT/5, EXT/15, etc. Note that you can't extend the flaps without the slats. The slats are quite necessary on the Nine. Landing without the slats increases Vref by 25 knots. Landing without slats and flaps increases Vref by 35 knots. That's around 170 knots at heavy weights. I'm not sure if that answers your questions. Let me know if you need more info. I'm always glad to help! Bruce Artwick and company wrote some really robust code back in the early 1980's I never realized how good it can be until I found the HJG models, the CalClassic models, and many others. Who knew? ;D All of you do great work, and I definitely appreciate it.
|
|
|
Post by Mike Monce - HJG on Mar 26, 2013 20:32:21 GMT
That corresponds well with what is in the cfg file. As for your other problem try leaving the pitch scalar alone and try lowering the lift scalar in 0.1 steps. I looked at the air file and the pitch scalar is multiplying a fairly small parameter compared to the lift scalar. That's why you had to run it to such a large value, with bad results. As it stands now the slats are actually providing some lift to the model, so be aware that changes there may need a corresponding change in the trailing edge flaps.... typical for this sort of FDE tuning Mike
|
|
|
Post by acourt on Mar 26, 2013 22:49:29 GMT
Mike, Thanks! I'll give that a try. I don't know how you guys do this FDE thing. Maybe I should just stick with classic scenery design! ******* I just tried some of the changes you mentioned, and while they got close, the airplane still pitched the wrong way. I'll keep working on that end, but I thought of another idea: I changed the slats' pitch scalar to -10.0, and changed the flap pitch scalar to +10.0, leaving all other parameters as you guys had them. The body angles still work out right, and it fixed the trim problem. In fact, the trim is perhaps a little too far nose down with the airplane in landing configuration. I think if I reduce the flap pitch scalar slowly back toward 0.0, I might find the sweet spot. Does this sound like a good fix, or am I possibly barking up the wrong scalar? When you're working on flight dynamics, how much ends up being "I don't know why that worked, but it did," and how much is rhyme and reason?
|
|
|
Post by aerofoto - HJG Admin on Mar 27, 2013 5:11:51 GMT
WELL .... ;D FS is full of "BULLnuts & JELLY BEANS" .... and we do, often, have to apply a bit of our own "BULLnuts & JELLY BEANS" .... in order try'n address certain issues .... and just hope, like hell, we don't "schucks-UP" anything else in the process ;D To put it all another way perhaps .... Not every FS issue can be resolved in accordance with absolute realworld fidelity and without risking upsetting something else .... mostly because of the limited number of paramters available within FS and which may be manipulated. It's often a very delicate dance balance between fact and fiction that results in various work-arounds and compromises nneding to be applied .... as both Mike and I (and probably every other FDE writer too) have learned in the past. Mark C AKL/NZ
|
|
|
Post by Mike Monce - HJG on Mar 27, 2013 15:04:26 GMT
This may be one of those cases where air file editing is also required to achieve what you want. I'll have to look when I get home, but the lift curves for the flaps/slats may be where things need be be more fine tuned. Ditto what Mark says, though ;D Mike
|
|
|
Post by acourt on Mar 27, 2013 15:36:44 GMT
Mark and Mike, "Bullnuts and Jellybeans?" I always love the phrases you come up with, Mark, and I really have to remember that one! Is that another "NZ-ism" or a "Mark-ism?" I understand what you're both saying. I've fiddled a little more with my project here, with results that, while not "spot on," get the point across. I think it's as close as I'm going to get with my limited FDE knowledge and few editable parameters in the .cfg file without fouling up something else. I'll make some last changes today, then let you try it so you can see the effect I'm talking about.
|
|