Stall Characteristics of the DHC-6 Twin Otter

[louross]

Here's what I have:



After opening your email a few days ago, I opened the two files to see what they said. The icon changed to this blue icon. Maybe that's the problem.

But, no a/c listed in FS9.
lr.

[Tom Goodrick]

I hope the forld called ORIG_Aircraft holds both your original aircraft.cfg file and your original DHC6W_22.air file. If it does, then delete those two other files we se and copy them back into the same place.

I'll send you two new and improved files. When they arive, delete those the aircraft.cfg and DHC6W_22.air files and copy in the two from the new email. You already have the originals saved so you can just delete the copies next time around.

Sorry but I still don't have the new .air file resolved.

[louross]

I did a flight in the BGA with the newest files you sent. I also moved the sound folder and the panel folder from the "Bradley" a/c from PAD into the BGA a/c folder and it looks and sounds much better. I will post a screen shot later of the correct cockpit set-up for take-off. That set-up also includes pilot/pax views.

(1) Here is the main problem. You should be able to slow to 80ias, flaps up or at 10*, and have pitch control. But you lose pitch control at that speed and as you descend and slow, the nose falls- or just maintains the pitch until you hit the ground. There is no ability to rotate or flare. You should be able to touch down at 60 (full flaps), but I have no pitch control. It appears you had the same problem.

(2) I noticed you eliminated the aux tanks- good! The fore and aft fuel tanks should hold equal fuel. Also- nice job on the pax loading! The panel I used never showed any fuel.

(3) Oil pressure reads too high, or is off the scale. Prop RPM should reduce to 75%, this shows about 83% or there abouts. There is no trim indicator- (the real a/c has a manual wheel on the right arm of the Capt's seat and if I remember correctly the indicator is also there). There is a warning flag on the No. 1 NAV.

Can these problems be fixed? It would then be a very nice plane for FS.
lr.

[louross]

You can noe see the cockpit setting for normal take-off and some instrument errors in the June screen shot photo.
lr.

[Tom Goodrick]

I can probably improve the lowspeed handling. RPM and oil pressure may be tough to change. We'll see.

As for the trim indicator, that has to be on your panel. It is on mine. I use a digital indicator and usually include a trim wheel from a Cessna for the analog indication. I'll hack away at these problems for a day or so. This is the first real data you've given me. It helps.

I just looked at your photos and read your remarks. Your fuel gauges are not looking at the correct tanks. The reason is that my files have the fuel in two center tanks. Your gauges are probaly looking at the main or main+aux wing tanks which should be empty because they do not exist as you have pointed out.

I'll look for some tank gauges that will work with my aircraft.cfg file.

I guess I don't understand what you find wrong with the torque and prop relation. I have found them to be independent because I usually leave the rpm constant and just manipulate the throttles (oops, sorry, torque levers). I see the torque change while the rpm stays fixed. I have a turboprop power gauge that I made that takes torque and rpm into consideration. It shows power as percent of max rated power based on readings of the percent torque and the percent rpm. (SHP% = N2% * TQ%)

I follow the standard Beechcraft King Air method of keeping the rpm at max during takeoff and climb while adjusting the torque to keep temps in line. Then at cruise I usually drop the rpms a bit while keeping torque near 80%. I leave the condition lever at low rpm. I have several books that describe this method. Some go into considerable detail.

Piper set up the turbprop engine on their new Meridian single so that the the rpm is locked at 2000 rpm. All you can do is manipulate the torque. I made a model of this and left the prop lever off the panel (But I show the rpm and torque. They certainly are independent in this case.)

[louross]

PT6A Engine- is a little different in certain respects.
(1) Fs does not model beta range of this engine- probably the basic code is not there in FS. If someone tells you it does, then they have no experience with this engine.
(2) B99; prop is det full forward (2200rpm for takeoff. Gear UP and reduce props to 2000 for climb. Cruise reduce props to 1900. Have no idea how the KingAir operates. DHC6; that power is not necessary, especially considering the noise. 75% is common, or up to 80%.
(3) Both B99 and DHC6 (and I would bet all Pt6A engines); Take-off climb power, or cruise power, and props set. If you reduceprop RPM, the torque will increase- if you increase prop RPM, the torque will decrease. If you take off with the props up, and reduce prop RPM while the torque is near the red line, you must reduce the torque or you will exceed red line. On a hot day excessive torque is not a problem because you will be temp limited (ITT gauge). On cooler days you can over torque if you're not paying attention (in pilot language this is called "flying with your head up your arse" ;D)
Remember- I put the "Bradley" panel in the BGA a/c.

[Tom Goodrick]

I ran a number of trials using a TPower gauge I wrote that shows throttle position in %, power %, thrust lbs, Torq %, porp RPM, and N1% (first stage rpm). Below 70% throttle, the RPM is mainly independent of the prop lever setting and varies with the throttle setting. Above 70% the prop RPM is constant and maxed out. The power and thrust varies with the throttle. These tests were all done on the ground with the aircraft held stiull with the brakes (to the extent possible).

I only checked prop settings with positive blade angles. Things do get interesting when you let the blade angle go negative with the throttle in the "positive" range (not the reverse thrust range).

[louross]

1) Fuel Gauges- calclassics redid their DC6B propgam a while back and had some problems with the fuel tanks/gauges. They were able to rename the tanks, like aux iinstead of main, or whatever. Maybe the same here.

2) PT6A Engine- real simple- power is up to xxx, props set. Increase prop RPM, torque decreases; reduce prop RPM, torque increases. Does this happen with other turbo props? No idea. Some turboprops have no prop lever- it's all handled with the power levers. Beta Range example- power levers at idle, prop levers full forward; push the power levers forward anywhere up to xxx amount (don't remember the setting, but rather lower power range) and prop RPM changes as at low power settings the prop RPM is contoled by the powers levers.

Typical procedure (airport- not bush):
DHC6- flaps 10*, prop lever to back and maybe then a titch forward to result in 75-80% prop RPM, power to torque red line (unless temp limited), rotate (about 70-75), accelerate and flaps up, establish climb (typically 120), maintain max torque to cruise altitude. Reduce torque as necessary for landing, props at the original 75-80% until on the ground then forward for reversing and taxi.

Beech99 (same engine- diffrerent airplane): flaps up, props full forward, max torque (or temp limited), rotate at 90, low nose attitude and gear up, accelerate to 120 as quickly as possible- low climb rate, increase pitch and props 2000 accelerating to 160, increasing power levers as necessary to maintain max torque possible, cruise set props at 1900, set cruise torque. Approach/landing (typical): cross FAF at 160 with approach flaps set (setting is about 35%), power reduced to about 400-450 lbs torque, begin slowing, about halfway down the glideslope and at about 140 lower landing gear. about 1/2-3/4 mile from runway at 120 and full flaps, reduce power and touchdown at about 85-90, props full forward and flaps up.

[Tom Goodrick]

You are right that an FS turboprop normally shows the higher rpm at lower throttle.

I just tried what I thought was a negative prop angle but it wasn't though it did give negative thrust.

With the power levers at 30%, setting the prop levers at 0 or 100% give 1133 rpm. But setting the prop levers at -16% give -88 lbs thrust/eng and 568 rpm. The blade angles went from 15 degrees to 79. degrees. In this case increasing the prop lever setting would increase the rpm. increasing the power lever to 60% would increase the negative thrust to -110 lbs and the RPM to 703. Blade angle remains maxed at 79.3 degrees.

Evidently this is how to get 'flat pitch" for added "drag' (=negative thrust) while taxing or during some steep descents.

[Tom Goodrick]

Here's what I know about fuel tanks. It is from info published by Microsoft:

Here are the tank names Microsoft provides that can be read by gauge. I can give you a digital guage that reads any of these tanks and we can define them to carry fuel in the aircraft.cfg file (also give them each a position).

FUEL TANK CENTER QUANTITY
FUEL TANK CENTER2 QUANTITY
FUEL TANK CENTER3 QUANTITY
FUEL TANK LEFT MAIN QUANTITY
FUEL TANK LEFT AUX QUANTITY
FUEL TANK LEFT TIP QUANTITY
FUEL TANK RIGHT MAIN QUANTITY
FUEL TANK RIGHT AUX QUANTITY
FUEL TANK RIGHT TIP QUANTITY
FUEL TANK EXTERNAL1 QUANTITY
FUEL TANK EXTERNAL2 QUANTITY
FUEL LEFT QUANTITY
FUEL RIGHT QUANTITY
FUEL TOTAL QUANTITY

Tank selectors can be made to feed from each of these tanks by number according to a list. There can be a cross feed valve but it is not clear between which tanks it works.

There is a normal default fuel use progression of tip, aux and main. Center tanks are used concurrently with tanks in the progression. I have not checked out all possibilities. I never have used external tanks. Don't know ehre they fit although I would hope they would have priority over internal tanks so they can be dropped.

I'll take a wild guess and say the number of tank selectors sets which selector feeds which engine. This is not clear. That way you would designate which engine is fed from which tank. We have this in piston twins using the fuel selector in the Baron. (It does not distinguish betwee main and aux tanks). There is also a cross feed between the left and right main tanks.

In flying piston twins, I often have to use the cross feed to balance the load because one side will burn more than the other.

I don't do art work as it is time consuming. But it is easy to make a digital gauge. It takes about 20 minutes to make it and test it. I have been tempted to try fuel selectors but have not done it yet.

[Tom Goodrick]

OK. I finally figured out that the condition I mentioned above where the blade angle gets big is FEATHERING! Duh!

But I tried something and it has very interesting results. I set the min beta in the aircraft.cfg [Propeller] section to -2.0 degrees.At idle, the blade sits at +1.0 degrees with prop levers forward and we get a very small negative thrust and RPM of 970. push the power lever forward to 6% and we see a blade angle of -2.0 degrees and negative thrust of -196 lbs/eng with still 970 RPM. push the power lever to 30% and you see -486 lbs thrust per engine with the blade still at -2.0. To make a takeoff from this condition, push the power lever to 70% and the RPM goes to 1700 and the thrust goes to 2109 lbs/eng with a blade angle of 18.3 degrees.

Is this a change in a reasonable direction? I have never tried a negative min beta before. I just started messing with this and there is more control of speed during taxi with either prop lever set part forward and varying power or power set and varying prop setting.

I'll send you this set of FD so you can see if it is worth any more work. I improved the stall characteristics a little.

[louross]

I hope this is what is needed here (don't remember the prop angles- it's been too long):
Prop levers full forward, engine idle: aircraft will taxi (add some power to get started, then idle).

Prop levers moved to the rear, engine idle: a/c won't move. I never tried to taxi with the prop levers full rear. Doesn't make sense to, would take too much power- noise, high temps, extra fuel.

Move prop levers from rear position to FEATHER: if taxiing, a/c will slow down to stop. If in the air, find a new job, assuming you are alive.

Taxi speed is contolled with power levers, prop levers full forward. Moving from a low idle setting to idle, to just barely into reverse, contols the speed. When you first initiate the power levers into reverse, the props will just barely go into reverse and the a/c will slow- when this happens, there is a very distintive sound made by the propellers.
To give an idea of some of the minute changes here is a working idea: at the gate everything's ready- brakes are released; props moved out of feather to full forward; power lever pushed up a bit- rpm increases- a/c begins taxiing; pulling away from the gate speed builds to a sufficient point- power levers to idle- a/c continues moving; the need to slow down arrives- (power levers are at idle) you "twist" the power lever handles "forward" and the props begin to enter reverse, the unique sound is heard, the a/c slows, if reverse thrust is actually needed, then at this point the power levers are pulled further back into the reverse range. During the entire taxi process, the power levers are moved forward or back controlling prop rpm- you don't need to push the power levers forward enough to actually increase torque.

Point: the engines are started and shut-down with the props in the feathered position.

As I remember, the fore and aft fuel tank has a line that feeds a manifold where the engines draw fuel from. Tank to tank is not possible.

MISC:
I am sure, but don't really know, that some of what we are talking about cannot be programed in MSFS. Also, it is my understanding that FS has too much wheel/runway friction, which is the reason a/c in FS don't taxi at idle and also that using actual a/c performance charts are invalid when computing balanced field lenths and V1 speeds.

FS programed fuel usage is not correct, except for certain a/c, as it is also my understanding that FS uses the C172 as the fuel model. On this point the correct, actual, fuel cfg method for heavy a/c is: burn center tank(s) fuel first. Then burn inboard main tanks to a point equal with outboard main tanks. When the outboard main tanks are below full, open the aux fuel valve to feed the outboard mains as the in/outboard wing tanks burn. Variances with the particular a/c ( eg 727, 767, 744) will be obvious once the fuel system of that a/c is understood. Some payware developers re-wrote the code in order to have their a/c use fuel correctly.
In fact, on a different note, PMDG actually re-wrote the weight/balance code in order to get the actual CG base for their 744 rather than the one that MS made up in the MS programming code.
lr.

[Tom Goodrick]

Some of what you say about the FS problems is correct and some is not. I hope to surprise you about some things. For example I know that few aircraft you can download will fly with correct fuel consumption but all of mine do. I take pride in setting mine up so that, if nothing else, you can fly a route in one of my aircraft (or set of FD fiels) and see very realistic fuel use.

After I get an airplane doing everything else right, I fix the fuel flow scalar at a known cruise condition (altitude, speed and power setting). The fuel consumption will generally be close at all altitudes and at other power settings. The C172 has nothing to do with it. I have done everything from a Piper Cub to a 747 and C5A over the years and never had a problem getting fuel consumption close.

I have not tried that yet with the DHC-6 because there's a lot of work to do and I don't yet have a test condition to match. This aircraft falls into a hole in my library.

Wheel friction is a big cunnundrum. Rolling friction is difficult to get right. I often lift the aircraft up enough so the friction is low but people complain they can see sunlight under some tires at some points while taxiing. A common problem is an undetected scrape point that is too low. But on the othere hand there is very little side friction. You can land crabbed 30 degrees into a strong wind and not tip at all. I have guages that calculate V2 for all the jets and I can make those pretty close after finishing all the other aspects of design. There may be something off in the rolling distance to V2 but it is not so much that runways that should be of use cannot be used. Stopping after a cut at V1 would be close in some cases.

CG is something else I get pretty close when it is important and good data is available. A few years ago, a pilot for a CitationJet was visiting this Forum and he gave some accurate data on loadings and CG. It took some work but there certainly was not any need to "rewrite the code". I nailed it. A few years before that there was a pilot of Learjet 45's that visited. He gave me a copy of his weight and balance computer program from his lap top. I matched that without a problem. It is not hard to do but the more accurate data I have, the more accurate I can make the result. Ussually all I can do is try for something near 15-20%mac at forward loadings and not much farther aft than 33% with an aft loading. When I did the Bonanza V35B, I was careful to make it get into trouble with a full house on a long flight. The CG moves aft with empty tanbks making some landings touchy.

I have instantaneous weight and CG position displayed on all panels. I care more about getting things all the aerodynamics and flight dynamics right than about having a "realistic" panel.

I am curious to hear what you think about this new set of FD's. Some of those characteristics you mention about taxiing are things I saw this afternoon. This may at least be a step in the right direction. Some fine tuning may be needed.

[louross]

Very interesting. PMDG said that MS places the datum line at the nose, whereas it should be at some point well in front of the nose. I was trying to remember what the datum was for the 727, but, again, it was too long ago.
BTW, check your email.
lr.
going to shut down in about ten minutes, too tired.

[Tom Goodrick]

The datum line just defines one system of coordinates. Airplanes often have several during their lifetimes. There is a basic longitudinal axis along which positions are measured. If no offset is specified, then FS assumes the zero point is on the longitudinal axis even with the leading edge of the mean chord. In earlier versions it was assummed to be at the 25% mac point of the mac. Aerodynamicists always refer to CG positions as a percentage of the mean aerodynamic chord (mac) starting from the leading edge of the chord. This is because stability equations predict various behaviors based on position measured that way. 5% to 30% is usually manageable with 20-25% being ideal.

To make loading problems from real aircraft work in FS, Microsoft provides an offset from a datum point near or ahead of the nose. You see this as the ref point x,y,z in the station list in the aircraft.cfg file. (Read as a text file.)

We start by placing the empty weight somewhere on the x axis, usually close to zero if no offset. Then we place the heavy things like engines and fuel tanks. Then we place the stations for passengers and cargo. (Crew is usually considered part of the empty weight except in small planes.) Then i look at the CG gauge showing the CG with pax and fuel for a typical flight. If the load is supposedly balanced, I will look to see a number near 20 to 25%. (I usually start with a balanced load.) If the gauge indicates a CG far from this mark, I move the empty weight a little bit and re-check. When this looks good I'll look at reasonable but extreme loading situations. Eventually I get a reasonable range of CG. I flight-check the extremes to see if I have exceeded a limit. Without hard real data, I have to base the range soley on observed stability. With real data we can usually nail it.