Tips and tricks from an idiot savant

[Heretic]

Still looking for decent DC-9-30 drawings.

The 6437 x 2638 version of this drawing will not download.
https://encrypted-tbn0.gstatic.com/...5JXN9ZxXOQ918GUKyiEbqtLKqflZV9rkyVtkrP0iSR3aA

Do you or anyone else have this drawing in its full size?

That link and image yields practically nothing. How did you find it in the first place?

The drawings in the DC-9 Airport Planning document look reasonably accurate:
http://www.boeing.com/resources/boeingdotcom/commercial/airports/acaps/dc9.pdf

Other than that, you can try to contact Peter, who runs MD-80.com. He should have some info.
https://www.md-80.com/impressum/

The "real thing" will DEFINITELY not reach 350 at a heavy weight. You have to "step climb." Does the climb rate 'fall off' at all as altitude increases after takeoff? If not, then generated lift is not decreasing as air pressure decreases with altitude.

What is atmospheric pressure and angle of attack at 350? AoA should be in the stall regime.

Basic flight characteristics aren't broken as climb rate still normally decreases with altitude. Just not enough to be believable.
I just took a look at the .air file tables for lift and drag. They match that of the default FSX 737-800, which is not the best donor. Even the tables for the 727 I did a while ago would be a better fit.
Fixing the tables isn't impossible. Convert the polars I got from Erick into an .air file friendly format and implement the engine tables from the 727, then test, test and test again. But it will take quite a bit of time.

 

[EdMcCauley]

Basic flight characteristics aren't broken as climb rate still normally decreases with altitude. Just not enough to be believable.
I just took a look at the .air file tables for lift and drag. They match that of the default FSX 737-800, which is not the best donor. Even the tables for the 727 I did a while ago would be a better fit.
Fixing the tables isn't impossible. Convert the polars I got from Erick into an .air file friendly format and implement the engine tables from the 727, then test, test and test again. But it will take quite a bit of time.

Another possibility is engine thrust does not diminish with altitude increase as it should.

 

[Heretic]

Another possibility is engine thrust does not diminish with altitude increase as it should.

Engine parameters are okay.

 

[EdMcCauley]

Does your model stall at a reasonable speed at low altitudes? If so, that would be a good indication that the angle of attack function is correct. If it does not, then your aircraft would continue to climb at high altitude despite correct air density and engine thrust.

BTW, is there a better place on the forum, a new thread for example, to post a plea for DC-9-30 wing "blueprints", airfoil thickness v. semispan, and geometric not aerodynamic twist? Some of this data is available for the -10, but the -20 through -50 series wings are quite different from the -10. Due to the above factors, a -30 sim model should have significant 'handling' characteristics than the -10.

A number of years ago I 'flew' a DC-9 on FSX. This was on a friends computer, so I know nothing about the DC-9 model used. If you want your model to fly like a DC-9/MD-80 not a generic airliner, it should respond MUCH quicker to aileron inputs than the 727, for example. The 9's ailerons do not use hydraulic pressure for movement so large inputs are required. In the takeoff and landing speed range, particularly on windy days, to maintain a heading required rapid, near full control input which must be returned to near neutral just as rapidly to avoid overcontrolling.

The DC-9 easily wanders off heading during an ILS approach if you don't watch it closely, but it can be returned to 'on course' just as quickly. This is important when dramatic shifts in wind velocity and direction occur. If you liked this characteristic, you deemed the airplane "responsive." If you did not like it, then you called it "unstable" or even "squirelly."

To continue the 727 comparison, this airplane, in landing configuration, was slower to respond than the 9 when a course adjustment was necessary, but once it was on course it stayed there. If you liked this, then you said it was "stable." If you didn't then it was "unresponsive."

Personally, I liked BOTH airplanes. They were just different.

 

[ErickC]

Remember that the designation DC-9-10 did not actually exist until the -30 came out (sort of like how there was no such thing as a 727-100 until the 727-200, thus the early two-digit customer designations like 727-25). If there's no suffix, assume it's the 10 (actually 11, 12, 14, or 15). The reason I made a note about the -30 wing on the diagram is because I was planning where the future cross-sections for the -30 wing would go, meaning my -10 wing has to take into account the location of the -30s slat boundaries so I don't have to do too much work to make a -30 wing (Douglas didn't need to worry about polygon edges and the resulting locations of flex points).

I might have misread the zero-lift line graph as a simple incidence of camber line graph. Oops. Still came out pretty close to the photos I compared it to, though.

EDIT: actually, all I have to do to figure out if it's a -10 wing or not is calculate the percent of span where the vortilon is on the -10 and -30 and see which one is closer. I'm pretty sure I already did that and the graph is for the -10.

 

[Roy Holmes]

Ed,
A few points.
If its flight dynamics that interest you, you might get more info from that forum. Not all the flight dynamics munchkins wander in here.

The incidence on swept-wing airplanes is usually more negative at the tips than the root to prevent the tips from stalling first. There is a span-wise component of airflow on swept wings that can cause tip stalling if there are no wing fences or wash-out (reduced incidence)

None of the stock airliners in FSX or P3D have the correct reduction of thrust with altitude because the engine tables 1502 to 1507 in their air files are not constructed the way that the sim software requires.

Are your comparisons in post 124 about your experiences in FSX or some real world data? My understanding was that the 727 was very responsive on the approach. However I'm almost completely ignorant about airliner performance, fast jets is my speciality
Roy

 

[EdMcCauley]

If its flight dynamics that interest you, you might get more info from that forum. Not all the flight dynamics munchkins wander in here.

Hello Roy, thanks for responding.

I'm looking for specific data on the DC-9. Wing parameters as stated in a previous post.

"Heretic" is the one asking why the DC-9 model in FSX is outperforming the real aircraft. I'm just suggesting possible reasons why this is occurring. Or are you stating that I would do better to post my data requests in the "Flight Dynamics Forum"?

None of the stock airliners in FSX or P3D have the correct reduction of thrust with altitude because the engine tables 1502 to 1507 in their air files are not constructed the way that the sim software requires

Assuming "Heretic" is using one of the "stock airliners" it appears you have directed him to the source of his problem.

Are your comparisons in post 124 about your experiences in FSX or some real world data? My understanding was that the 727 was very responsive on the approach. However I'm almost completely ignorant about airliner performance,

The DC-9 was the first airliner I ever flew, and the first one I checked out on as a captain. Subsequently, I flew all the Boeings from the 727 to the 767 for many years.

Once in landing configuration the 72 did not fly like a fighter...to put it politely.

fast jets is my speciality

You're a lucky guy. Which fighters have you flown?

 

[ErickC]

Oh, yeah, before I forget, I did, during the time I spent gathering references, create a good-enough-for-model-building collection of airfoil guides for the DC-9 based on diagrams from journal articles, maintenance manuals, and photos of torn-up wings (which made the cross-sections apparent). I compared them to photographs I took of a museum piece (unfortunately the exhibit was under construction, so I wasn't able to get terribly close) and they look about as close as I can get without either blueprints, measurements, or a 3D scan (which I'd love to do, maybe I ought to get a Kinect and see if Delta finished that exhibit yet). I'll see if I can dig them up for you.

 

[N2056]

Perhaps the thread title could be changed..."Planes, Trains, and DC-9s"
All in good fun. Fact is I'm a train nut myself. I even played hobo once and rode a freight train between Dunsmuir, CA and a bit south of Klamath Falls, OR.

 

[EdMcCauley]

Oh, yeah, before I forget, I did, during the time I spent gathering references, create a good-enough-for-model-building collection of airfoil guides for the DC-9 based on diagrams from journal articles, maintenance manuals, and photos of torn-up wings (which made the cross-sections apparent). I compared them to photographs I took of a museum piece (unfortunately the exhibit was under construction, so I wasn't able to get terribly close) and they look about as close as I can get without either blueprints, measurements, or a 3D scan (which I'd love to do, maybe I ought to get a Kinect and see if Delta finished that exhibit yet). I'll see if I can dig them up for you.

ErickC,

I would greatly appreciate anything you can come up with!

Have you tried going direct to Boeing for DC-9 blueprints? After 50 years maybe they let loose of it...

 

[Roy Holmes]

Ed,
http://www.dept.aoe.vt.edu/~mason/Mason_f/B717PresS07.pdf
That might be useful.
The DC-9 series airfoils are DSMA - Douglas Santa Monica Airfoils.
Douglas DC-9-10 DSMA-357/-358 DSMA-359
Douglas DC-9-20 DSMA-433A/-434A DSMA-435A/-436A
The middle column is for the root and the right hand one is for tips.
All the models after the -20 (up to the -80) have the same airfoils.
Now, finding coordinates for those airfoils is tough because they are most likely proprietary.

The actual shape of the airfoil and how it is modeled has no effect on the sim aerodynamics. What is used is primarily a fuselage AOA versus CL curve. and there is one in the PDF for the wing with flap settings of zero to 50. It says it is "similar" to the DC version.
The shapes shown in the PDF look odd to me. The Root one is most likely the section where the wing and fuselage meet so that shape is for the wing/fuselage fairing at that point and a lot "fatter" than the root section would be. The tip section could be accurate for a washed out airfoil. The drag coefficients given are probably about 1/3 of what they should be.

Heretic (Bjorn) and I have had many discussions over the years and he has used the data on engine performance I put in a paper as a basis for his models.

My fighter time was 1800 hours (9 years) on the Hawker Hunter. and 1200 hours (6 years) on the F-4M. The background airplane in my mug shot is a TA-7C that I flew at Patuxent River on a visit there.
These days I fly at least 4 times a week with virtual squadrons and I usually fly the Tornado. It is the closest modern airplane to the mud movers I flew as a lad. My hobby is flight dynamics development and I have done those for 9 fighter models which are published (not by me) and have 2 more in the works.

My point about the Flight Dynamics forum is that the folks who frequent that forum have more clue on that subject than the model makers/painters here. And before I get polygons and pixels fired in my direction I will hastily state that the opposite is even more true.
Roy

 

[Heretic]

Does your model stall at a reasonable speed at low altitudes? If so, that would be a good indication that the angle of attack function is correct. If it does not, then your aircraft would continue to climb at high altitude despite correct air density and engine thrust.

Thanks for the continued questioning, but it doesn't dig into new teritory for me. Lift, drag and engine tables need a complete overhaul and that's it. Stall caharacteristics are part of that.

The DC-9 easily wanders off heading during an ILS approach if you don't watch it closely, but it can be returned to 'on course' just as quickly. This is important when dramatic shifts in wind velocity and direction occur. If you liked this characteristic, you deemed the airplane "responsive." If you did not like it, then you called it "unstable" or even "squirelly."

To continue the 727 comparison, this airplane, in landing configuration, was slower to respond than the 9 when a course adjustment was necessary, but once it was on course it stayed there. If you liked this, then you said it was "stable." If you didn't then it was "unresponsive."

Personally, I liked BOTH airplanes. They were just different.

Both behaviors can't be exactly replicated in flight simulation because the fidelity of the real thing is not there. You can only give people an idea of how these aircraft fly.

Assuming "Heretic" is using one of the "stock airliners" it appears you have directed him to the source of his problem.

No, it's a payware model and not stock at all. The devs did put some thought into the flight model, but didn't go the extra mile.

The DC-9 was the first airliner I ever flew, and the first one I checked out on as a captain. Subsequently, I flew all the Boeings from the 727 to the 767 for many years.

DC-9s, 727s and 767s. Sounds like Allegheny and then USAir/US Airways.

Once in landing configuration the 72 did not fly like a fighter...to put it politely

With flaps 30 and low throttle, it didn't fly at all. What was your ratio of greasers to tooth rattlers in the 727?

 

[EdMcCauley]

Thanks for the continued questioning, but it doesn't dig into new teritory for me. Lift, drag and engine tables need a complete overhaul and that's it. Stall caharacteristics are part of that.

Let the FUN begin! Good luck.

What was your ratio of greasers to tooth rattlers in the 727?

It wasn't as tough to land as you hear. IMO it got a bad rap because it was the first jet for so many long-time prop pilots. The senior guys got the jets first, and they weren't properly taught how to fly a jet. The longer you had flown props the more deeply ingrained your habits were. Many habits had to be unlearned.

High sink rates at low altitudes were a major problem in the early days of the 727. Unspooled jet engines (especially first generation turbojets) take a while to begin to produce power, and even worse there's no propeller blast of air across the wing to immediately produce lift to stop a high sink rate. In a jet you have to wait until speed increases to acquire more lift. Thrust vs. mass and inertia are the issues.

A stabilized approach is the key. Coming across the threshold with a high or increasing sink rate or multiple gross power changes is asking for a hard landing. With aft-mounted engines your center of gravity location is also aft. If you pitch up to counter a sink rate you actually are pushing the main gear rapidly downward. Add this to the existing high sink rate, and you get a harder landing than if you had done nothing.

In windy conditions, the trick was to keep the descent rate under control with power and make small attitude changes during the flare. This worked just fine in the DC-9/MD-80 and I didn't have any landing problems transitioning to the 727. IMO, all three are in the same degree of difficulty category.

You fly with a different crew every month; which means I personally only made half the landings. So, it's not just me bragging. Not every landing was a greaser, but a "tooth rattler" was a rare event indeed.

 

[ErickC]

All right, a few things to keep in mind:

1.) The airfoil images are composites generated from several sources
2.) They seem to mirror reality fairly closely, but there can be no guarantees

The first image is the DC-9-10 airfoils and is a composite of figures from the maintenance manual and from a journal article that escapes me. The annotations are geared towards building 3D wing geometry and are mostly useless, but do indicate span percent and incidence (taken from the graph, may not be correct, as was pointed out, the graph is reporting the incidence of the zero lift line). Note that I drew a line from the apex of the airfoil to the trailing edge, and then rotated the images until this line was straight, so that I could build the geometry by the numbers. So consider these drawings to be at zero degrees incidence.

The second image is from the DC-9-10 maintenance manual. The red outline is the -30 leading edge. I superimposed a photo of an Aeromexico DC-9 of Playboy heritage that is now a static display. The wings were sliced just outboard of the fuselage, and the slats were not present. This exposed the flat spot on the wing body fairing where the slats cover up the leading edge. It was simply a matter of tracing the outline, and positioning it on the drawing until it fit (double checking with thickness to chord approximations based on photos and the like).

This is just the root... but the thing about MacDac's civil division, as much as I love 'em, is that they were perpetually short on cash and tended to resort to quick fixes (Proof? The DC-9-80: an airliner built to meet the needs of 1971, but delivered in 1979. Further proof? The "let's forget about the aerodynamic problem today and design a new fairing tomorrow" ideology). I have a hard time believing that the slat airfoil changes much beyond being squashed to fit the thinner wingtip while maintaining roughly the same chord. The 737s doesn't, and Boeing had money. So you could extrapolate the slat airfoil to the tip, squash it to fit, keeping in mind the proper locations of the slat edge, and it'll probably be close.

Here are some more Douglas diagrams I scoured off the almighty Google:

The above image was the starting point for my higher-resolution composite.

The original numbers were kind of hard to read, so I replaced them with more legible text.

Remember that the core wing is the same on all DC-9s, even the MD80 wing core is the same, just with a tip and root extension. So, you can compare this to a -10 station diagram and figure out fairly precisely where the -10 wing has to be modified to make a -30 wing (I say -30 because the correct way to count in DC-9-ish is 10, 31/32, 40, 20, 50, 33/34).

DC-9-10 wing stations:

DC-9-all vstab stations (don't trust the tailcone contours, the DC-9 cone points up a bit):

DC-9-all hstab stations (note that you get the vstab airfoil in this image and the hstab airfoil in the last one):

I am not totally certain how the raked wingtip fits on the -10... but it's not in any of the figures, so I think we can ignore it when building the core wing, and just consider it an aft skin extension (i.e. don't worry about how it fits into the tip airfoil).

So this should be enough to build either a -10 or -30 wing. I decided to go chronologically, so I built the -10 wing first, and, someday, if the model is ever finished, the -30 modifications will be done appropriately.

 

[EdMcCauley]

So this should be enough to build either a -10 or -30 wing. I decided to go chronologically, so I built the -10 wing first, and, someday, if the model is ever finished, the -30 modifications will be done appropriately.

Thanks for sharing all this!

I'll have a house full of relatives until new year, so I doubt I'll have any time to analyze all your data. But I will DEFINITELY get to it ASAP!

Do you have any info on the max camber and it's location?

 

[antaris]

Just a quick heads-up, although only applicable to FS9. I like to use alpha channels to control transparency because it allows all kinds of interesting effects that you can't get with transparent textures. I set up the transparent material here and this was the result:

Hello Mr. Cantu
Would you be so kind to show me the properties for this neat glass in Gmax? I just can't get it right (I forgot to mention that is to be used in FS9 as well)

Here are my settings within Gmax for the material:

I am using one texture; the setting within Gmax is the following:

Finally, the texture itself:

In advance, thank you kindly.
Sergio Kauffman.

 

[tgibson]

For glass that works in FS9 and FSX, I set the Opacity to 99% (not 100% because GMAX will not handle it properly), then I use the alpha channel in the texture itself (in the actual FS texture) to control the amount of transparency desired (black is fully transparent).

Hope this helps,

 

[antaris]

For glass that works in FS9 and FSX, I set the Opacity to 99% (not 100% because GMAX will not handle it properly), then I use the alpha channel in the texture itself (in the actual FS texture) to control the amount of transparency desired (black is fully transparent).

Hello Tom,
I will try it! Thank you very much.

Sergio

 

[ErickC]

Sorry, I've been busy with school and Open Rails-related projects (namely, the payware Soo Line GP7, and the creation of an all-new sonic environment to be put into the public domain, because the default 11kHz MSTS sounds that have been grandfathered into OR are all garbage). In FS9, your material must not have an environment map specified (I think that's the specular channel). Then, you control opacity with the diffuse alpha, but beware the position of the parts in the hierarchy, as this will determine draw order (I think I mentioned that a few pages back). In FSX, I think the option is "blend diffuse by diffuse alpha," and I think that you must then use the alpha channel of the specular map to determine environment map blending (which probably makes the most sense anyway, I think the specular alpha controls the sharpness of the highlight, and materials that have sharper highlights do so because they are smoother and thus will tend to be more reflective). Both of these are options that must be checked in the material editor, as far as I know (but my FSX experience is somewhat lacking, as I stopped building airplanes very early in the FSX days, and haven't really gotten back up to speed yet). I think there is something that needs to be done with the alpha test options, but you'd have to ask someone else about that.

The only game I've made content for that intrinsically handles alpha sorting the right way - that is, automatically - is Open Rails. There just ain't no reason to make people jump through all those hoops on the off chance you might want to hide something behind another layer (which is what you can do with alpha sorting). But ACES has always been fond of doing things the hard way, so that's what you deal with in MSFS.

 

[ErickC]

So I finally bought a monitor for the computer I bought off my sister's boyfriend when he upgraded, which meant migrating everything over to a Windows 10 system. Once I turned off all the crap I don't care about, unpinned everything I'll never use, and installed a program to prevent the computer from restarting without my approval, I found out that it's a perfectly fine OS and GMax works without issues... except for one. I may have mentioned that I map by applying a planar map of a defined cubic size (temporarily using "box" parameters to set the "W" size), then adding edit mesh and unwrap UVW modifiers to move and rotate things where I want them, before adding a final unwrap UVW over the whole mesh to flatten the coordinates and weld anything weldable. But, when I made the switch, I discovered that the mouse wouldn't zoom in the unwrap UVW window! This is a huge time saver, so I was a bit upset. But there is a solution. Open your device settings, click mouse, and uncheck "Scroll inactive windows when I hover over them." Problem solved!