Twin Otter X: open issues

[pbearsailor 9]

I´m not sure about that reducing prop rpm should increase torque!

When You retard the prop levers, You "tell" the props to rotate at a certain slower speed. To do so the propeller blade angle will be increased, wich will "bug" down propller rpm until this certain rpm is reached. The gasgenerator will just continue producing the same amount of air to drive the free turbine, cause You didn´t add more fuel for increased power.

Regards

Finn

Torque does increase with prop rpm reduction and it is necessary to reduce power with the power levers before prop rpm reduction when running at torque redline.

Pardon the example from another airplane (but same basic engine). The FAA requuired the early model King Air to have an electric backup system to keep the prop from going into reverse pitch in flight, called a secondary low pitch stop. It was awful and IT failed at times and that caused the prop to go into feather in flight. If it did it while cruising at power (which was common), the prop started to go to feather and the torque would begin to jump off the gauge if you didn't immediately retard the affected power lever. Most guys had a nylon tie wrap attached to the offending circuit breaker so you could find it quickly, pull it, and disable the system. That gave you your prop control back. After a few of these failures you got quick enough with it that nobody even noticed what was happening. :shock: Early Twin Otters have a similar system, but I don't remember many failures.

cheers,

steve

[Ken Stallings 1]

I´m not sure about that reducing prop rpm should increase torque!

When You retard the prop levers, You "tell" the props to rotate at a certain slower speed. To do so the propeller blade angle will be increased, wich will "bug" down propller rpm until this certain rpm is reached. The gasgenerator will just continue producing the same amount of air to drive the free turbine, cause You didn´t add more fuel for increased power.

What I think is "mixed" into this discussion is the way free turbines are used in helicopters. On helicopters with a free turbine, You have a governor that actually will add power to counter the drop in rpm when the blade angles on Your rotor (say propeller) is increased, thus requiring more torque to MAINTAIN rotor rpm (say propeller rpm).

The main reason constant speed propellers are used is due to the fact that a fixed propeller only is effective at certain RPM´s. A propeller with it´s blades fixed at fully fine would only be effective for take off and short slow flights. A propeller with it´s blades at a course setting would be economic at cruise speeds, but next to useless during take off.

I know that the turbine model of the Aerosoft Twin Otter isn´t perfect, nor are any other addon turboprop aircraft. Things could be improved for sure, but the policy of Aerosofts internal developement is to stick to the FSX SDK.

You might change the airfile to reflect a more accurate behaviour, but You have to assure that this behaviour also is accurate at both sealevel, 10.000 ft, -20°C and +35°C. If not You haven´t made a more accurate engine model, just a different one.

And all those engine parameters also effect airspeeds at altitudes etc.

What You get with the Aerosoft Twin Otter, is an aircraft that is flying close to actual real world performance figures when it comes to stall speed, cruise speeds etc.

Regards

Finn

Finn,

I need to correct a basic aerodynamics issue in your post. The fixed pitch prop does not lose efficiency at various RPM's. Rather it loses efficiency as airspeed increases. This is because of a principle called effective pitch angle of the prop blades.

As the aircraft achieves forward velocity there is an airspeed going across the prop blades in the horizontal direction. As this horizontal velocity increases the prop blades lose effective pitch angle. It is complicated, but it is a matter of how much air the prop can cut through in a single arc. With the aircraft standing still, it cuts through more air than at a given blade angle vice with the aircraft going forward. To correct for this reduction a constant speed prop governor will rotate the blades flatter to get a larger bite of air (angle of attack increase).

This flattening of the blade compensates for the loss of effective pitch angle. A prop blade has two air velocities. One the result of the blades rotating through the air. But it has a second air velocity -- the velocity of air moving across the blades horizontally from aircraft forward airspeed.

The constant speed prop also allows the pilot to set a specified RPM value. This is done because the engineers determined that given RPM settings produced given results. Most complex aircraft have performance charts. These charts show manifold pressure (or torque or ITT) settings with associated RPM settings to achieve optimal thrust, or optimal fuel efficiency, or the optimal combination of thrust and fuel efficiency.

In fact, many pilots use their own settings based on personal experience and there are conflicting theories out there with some pilots singing the praises of setting the lowest possible RPM with the highest possible MAP or torque/ITT settings.

Personally, I just set what the manual says to set.

But the key point is that it is brake horsepower (thrust) that one would wish to maximize during critical phases of flight such as takeoff or trying to clear an obstacle. It is true that when a plane stands motionless on the runway you achieve maximum BHP if you put the props into the maximum RPM setting (prop condition lever full foward and blades at the least flat position in rotation).

However, when the airplane starts to roll and develop forward airspeed, actual BHP is reduced because of the effect of relative pitch angle. If the pilot wanted to extract every ounce of BHP he would actually need to make very small adjustments to the blade angles to flatten them out. Of course, this would be impractical during takeoff roll as the pilot must concentrate on other more important issues.

However, most checklists specify a reduction in manifold pressure and RPM settings after takeoff and after clearing immediate obstacles. In the Baron for example, the checklist says to set 25 inches of MP and 2500 RPM for the climb and then maintain Vy until clear of all terrain, then level to enroute climb speed of 105 KIAS for optimal engine cooling.

The same theory works for turboprop aircraft with the difference being instead of manifold pressure the pilot rates his power output by referencing torque and inlet turbine temperature (ITT). But the exact same effective prop blade pitch phenomenon takes place.

You have two things working with a constant speed prop. You have the governor making minute and automatic changes in blade angle to compensate for changes in air. You also have manual changes in blade angles by the pilot moving the prop condition lever. In the first situation, the governor is automatically making these adjustments to keep the RPM steady. In the second situation the pilot is essentially directing the props into a different RPM setting.

Think of it this way, and this is also a great illustration of the principle of losing BHP. Go up in an actual complex aircraft. Have the pilot fly at cruise settings, normally something like 23 inches of MP and 2400 RPM's. With the aircraft stabilized at cruise airspeed, ask the pilot to advance the prop lever to the full forward position.

As he moves the prop condition lever forward you will feel a substantial reduction in airspeed. In fact, done quickly it's like the pilot engaging an airspeed brake. In fact, he has! The props in the full RPM position are producing less horsepower due to the effective pitch angle. The effect you feel in the airplane is like the hand of God pushing you forward in the seat just like you'd feel in a car if the driver hit the brakes hard!

Cheers,

Ken

[Ken Stallings 1]

Torque does increase with prop rpm reduction and it is necessary to reduce power with the power levers before prop rpm reduction when running at torque redline.

Pardon the example from another airplane (but same basic engine). The FAA requuired the early model King Air to have an electric backup system to keep the prop from going into reverse pitch in flight, called a secondary low pitch stop. It was awful and IT failed at times and that caused the prop to go into feather in flight. If it did it while cruising at power (which was common), the prop started to go to feather and the torque would begin to jump off the gauge if you didn't immediately retard the affected power lever. Most guys had a nylon tie wrap attached to the offending circuit breaker so you could find it quickly, pull it, and disable the system. That gave you your prop control back. After a few of these failures you got quick enough with it that nobody even noticed what was happening. :shock: Early Twin Otters have a similar system, but I don't remember many failures.

cheers,

steve

You have the stage to yourself on this one since I have not yet flown a turboprop equipped aircraft, much less one without a direct connection between the engine drive shaft and the prop shaft. I was a navigator on C-130's but not a pilot.

However, what you say matches precisely the techniques I've been taught in my dual piston training.

Is it right to say that regardless of the mechanical linkage systems, that when you flatten the prop blades with a given power setting (measured in MP, torque, ITT) that the demands upon the engine increases due to the increased air resistance in the rotational arc?

The reason you reduce MP before reducing RPM is simply to avoid overstressing the engine. I presume the exact same is true for both piston and turbine powered prop aircraft?

Cheers,

Ken

[pbearsailor 9]

I presume the exact same is true for both piston and turbine powered prop aircraft?

Yup, just keeping them below the redline.

[Finn 873]

Thank You guys for those indepths explanations of how engines and propellers interact.

The beauty of this hobby is that I learn something new every day :-)

But to return to our twotter:

While Aerosoft has chosen to stay close to the FSX SDK, there is no way to have an turboprop model that is very accurate throughout it's entire performance range. And I doubt that it is possible without extensive programming "around" FSX's turboprop model.

Finn

[pbearsailor 9]

But to return to our twotter:

While Aerosoft has chosen to stay close to the FSX SDK, there is no way to have an turboprop model that is very accurate throughout it's entire performance range. And I doubt that it is possible without extensive programming "around" FSX's turboprop model.

Finn

For sure, Finn. And I think following the FSX SDK was a good decision giving us a simple model with nice frame rates. But she's a great airplane just like she is anyway and with the new flight model probably is as close as any plane in flight sim that I've flown in real life (and that's more than a couple )

cheers,

steve

[Finn 873]

The main problem for a better engine model is, that to make it react correctly through all performance ranges, custom programming is required.

This is what companies like PMDG and Level-D are doing on their 747-400 and 767-300.

If You, like Aerosoft, keeps straight to the SDK, there is no way You can make the turboprop behave 100% or even 95% correct for all flight regimes.

What has been the top priority for the flighmodel developer of the Twotter, is to make a flightmodel that makes the aircraft fly as close to realworld figures as possible. Inertia has been one of the most important things.

Though I can´t be regarded as objective, I still find that the Aerosoft twotter is the aircraft in my hangar (and believe me I have alot of addons) that gives me the best feeling of the actual size and weight of the aircraft. It is relative agile whilst I still feel a bit of "mushing" through the air.

We could alter the engine model to be accurate at cruise speed f.ex, but then it wouldn´t be accurate at max performance or near idle speeds. It would probably only be accurate at certain altitudes and suffer at others. And we would even risk that these altered perfomances would "destroy" our present flightmodel, when it comes to speed, climb performance etc.

Regards

Finn

[JAL 0]

Howdy Ken, Steve and Finn,

Yeah, sometimes making changes like these can open up a can of worms. Just for the fun of it I'm going to do a bit of flight testing based on figures from a Twin Otter manual and see how she does. I do have some ideas for tweaking aspects of the air file if need be. Editing flight dynamics is just a hobby of mine, so I may run up against a brick wall, but the fun is in the trying, eh?

I'm not sure whether you guys have edited air files before, but I'd be happy to let Ken and Steve know how to make the mods for the engine if you want to try it out. Or maybe I can just e-mail you the modified air file? With private messages disabled here its kind of tough to exchange e-mails but maybe aerosoft can somehow pass mine on to you two if you'd like.

At any rate I will keep you posted here on this board.

Cheers,

John

[JAL 0]

I just did a quick test at sea level and found great similarities in the numbers between aerosofts model and the one with more accurate engine simulation. Fuel flow is a bit off for both sim models versus the manual. Sim FF is too low at low power settings and too high at high power. It would all average out though over the course of a flight, so its not worrisome.

Tests were performed in standard atmosphere with the twotter weighing in at 9000 lbs.

Setting 1

Prop Rpm: 96

Torque: 50

FF: 4.5/4.8/4.1 (AEROSOFT/ACCURATE ENGINE/TWIN OTTER MANUAL)

IAS:175/175/174

Setting 2

Prop Rpm: 85

Torque: 50

FF: 3.8/3.8/3.7

IAS:164/163/166

Setting 3

Prop Rpm: 75

Torque: 40

FF: 2.8/2.8/3.1

IAS:147/146/148

Setting 4

Prop Rpm: 75

Torque: 20

FF: 1.6/1.6/2.1

IAS:113/113/112

I didn't try at different altitudes or temps but I doubt there would be any difference between the two sim models. In setting 1 the accurate engine fuel flow is higher than aerosoft's, but I have an idea how I might fix this.

One other minor issue I see is that sound files are not done correctly in that the sound of the prop is linked to the torque gauge and not the prop gauge. Unfortunately pretty near all FS turboprops are like this. In fact I have only ever found one sound package that does this correctly and it is a specially designed freeware sound package for the King Air.

Anyway, this is just entertainment for me (and actually a fun way to learn) so it doesn't matter to me if aerosoft adapts these changes or not. It's been great to discuss this here on the board with such knowledgeable folks.

John

[Ken Stallings 1]

I'm not an airfile kind of person. That's a black art for me.

In terms of accepting your airfiles, please don't take this the wrong way, but let me wait for the Aerosoft patch and see what they have done. Pbearsailer reports they really have nailed the engine and flight dynamics and that's pretty promising for me.

Cheers,

Ken

[pbearsailor 9]

Howdy Ken, Steve and Finn,

I'm not sure whether you guys have edited air files before, but I'd be happy to let Ken and Steve know how to make the mods for the engine if you want to try it out. Or maybe I can just e-mail you the modified air file? With private messages disabled here its kind of tough to exchange e-mails but maybe aerosoft can somehow pass mine on to you two if you'd like.

At any rate I will keep you posted here on this board.

Cheers,

John

Hey John,

I'd be up for continuing to tweak it for our own use for sure. I think the changes Aerosoft has made that I tested have made great improvements and I'm happy with it now and understand their desire to wrap it up and get the boxed version done. Having said that, though, I've got no problem doing some further stuff with it if we can continue to get it closer to the real thing. My email is my nick here at gmail.com.

cheers,

steve

[Finn 873]

Hi John, Ken and Steve

The engine changes will not be done to the 1.10 update.

We are close to have finished the update and will not risk puitting some "worms" into it.

The 1.10 update contains new .air files and I would be pleased if You John would edit the files and send them to me, cause I would be happy to test them. We might be making a 1.11 update later, but right now we will stick to what we got.

Regards

Finn

[JAL 0]

Hi again,

Sure! Like I say I'd be real happy to send the modified files and keep seeing if we can get even closer to the real thing.

Before I send any files though... Finn, I've done the engine modifications to the 1.02 air file (from the BA model). Is this the same air file that will appear in the 1.10 version? If so then I'll send my mod to you and Steve asap, otherwise I'll wait until the 1.10 version comes out and do the mods there and then send them to you.

Great to hear that you might consider a 1.11 version down the road! For example, I would probably need help with issues like the sticking torque gauge and modifying a sound file so it would be super if Aerosoft were on board.

Cheers,

John

[JAL 0]

Easier than I thought! Using the freeware King Air sound file as an example I managed to get the propeller sound to link to the prop gauge. It's still in its early stages of development (i.e. I'm not exactly sure how I managed to do this...) but it sounds great!

Will keep you posted,

John

[Finn 873]

Hi John

You must know that I am not Aerosoft. It is up to Mathjis to decide what will happen after 1.10.

I have just taken the freedom to help coordinate and fix the issues, found after release of the Twin Otter. Mathjis has been supporting me big time. He also want to assure that the twotter will please as many users as possible.

For the .air files, please wait, cause they have been altered in 1.10.

From user to user, I have to say that You have to be carefull not to distribute those edited files, cause I don't think it is legal. Better ask Aerosoft first.

I'm sure that Aerosoft will appriciate if You can make their product better.

The problem is just that if users install Your files and are getting problems, we will have problems supporting them. Often people "forget" to tell that they have "tweaked" some files.

Regards

Finn

[Aerosoft Team [Inactive Account] 51558]

Hi John

You must know that I am not Aerosoft. It is up to Mathjis to decide what will happen after 1.10.

I have just taken the freedom to help coordinate and fix the issues, found after release of the Twin Otter. Mathjis has been supporting me big time. He also want to assure that the twotter will please as many users as possible.

For the .air files, please wait, cause they have been altered in 1.10.

From user to user, I have to say that You have to be carefull not to distribute those edited files, cause I don't think it is legal. Better ask Aerosoft first.

I'm sure that Aerosoft will appriciate if You can make their product better.

The problem is just that if users install Your files and are getting problems, we will have problems supporting them. Often people "forget" to tell that they have "tweaked" some files.

Regards

Finn

If we know about what is about to be shared it mostly is not a big issue for us. If thee content is something 'complete', like a full flight model we rather have the files distributed over our protected channels, but that's easy enough to handle. Just contact us on info@aerosoft-shop.com

[JAL 0]

Hi guys,

I'm just about to send a couple files to Steve and Finn for some testing. (Ken if you'd like to get them too just let me know.)

In addition to the engine fix it was necessary to work out a fix for the sound.

There was one additional problem in that at approach speeds and lower torque the prop rpm did not fluctuate with power inputs in the original aerosoft version. Instead once the props levers were pushed forward the RPM remained high throughout the approach. While looking at a few youtube videos I figured out this should not be the case:

e.g.,

SABA approach

http://www.youtube.com/watch?v=0JP-yvfs-U0&NR=1

KP160 approach

What should happen is that as IAS reduces (below about 86-90 knots) the prop rpm should start to come down when torque is also low. When the power levers and torque are increased then prop rpm should come back up. Effectively you should be able to control prop rpm with the power levers.

Anyway, this wasn't a simple problem to work out, but I seem to have found a way. So far in my testing I don't find any difference between the actual flight performance of the aerosoft model and the one with my mods. This is important because I think Steve is right that the aerosoft flight model is very good, and my modifications do not seem to affect it (let me know if you find something though!)

With the mods I think what we have is a simulation that is getting really very close to real life!

Still on my wish list would be to modify the gauges so that they do go over redlines... I think to do that I would have to buy some payware software, and at the moment I'm a bit reluctant. if anyone knows of some freeware software to make such mods let me know.

Cheers,

John

[JAL 0]

Hi again,

After doing some more detailed testing I came across something that differs between the two versions of the air file.

If you want to reproduce the tests that I've done you'll need a nifty program called AFSD, which gives you all sorts of detailed info about aerodynamics (among others) of the aircraft while it is flying in flight sim. It's a quick download at:

http://pagesperso-orange.fr/hsors/afsd.html

Steve I think you might be able to help us out here the most. Think back to an approach of about 74 knots or so with 20 degrees of flap. As you entered the flare and pulled back on the power levers could you 'feel' the props produce drag as the rpm slowed down? i.e., could you feel a braking effect of the props before you used reverse thrust? I do have a video of such a landing that might refresh your memory. It's not on youtube but I'll try and e-mail it to you.

In the aerosoft air file when you do this the props produce a significant amount of drag and the aircraft doesn't float much and it sits right down. This appears to be because in the original air file the props are still churning at very high (unrealistic) rpm.

In the modified file where prop RPM is correct, the props do not produce drag and the aircraft floats more in the flare.

In AFSD if you do these flight tests and look in the aerodynamics menu at the 'Net thrust' readout you'll see what I mean.

I guess the question here is whether propeller drag is modelled 100% correctly. Certainly the prop RPM in the original aerosoft version is incorrect for this airspeed and power setting, but the amount of drag produced in this phase of flight may be realistic. If so then it would be necessary to change how drag is modelled so it is also produced at lower prop RPMs.

I can modify my file to get a compromise between what it is now and the drag that aerosoft's file produces, but tackling how prop drag is modelled will require some more time and thought.

John

[Ken Stallings 1]

I realize that the engine changes were not planned for 1.10. However, just to make it official, the prop feathering issue remains. Only by chopping the fuel condition lever does the props feather in autofeather. The props cannot be feathered by a manual method.

Cheers,

Ken

[JAL 0]

Hi Ken,

My prop rpm is set up on the default keys I think. ctrl-f1 is reduce rpm all the way, ctrl-f4 is advance all the way. ctrl-f2 is reduce slowly, ctrl-f3 is advance slowly.

I was confused at first because when I pressed ctrl-f1 the props just remained at about 75%. BUT when I press ctrl-f2 over and over the props go to feather. This seems quite useful actually, because you can use the ctrl-f1 as a way to get the prop rpm to 75% but not go over into feather.

I'm not sure if this is where you had your problem but hopefully it helps.

cheers,

John

[Finn 873]

I´m not sure what You mean Ken.

If You think that the props should stop by only feathering the props You are wrong. The prop will run as long as the engine is running.

I can perfectly feather the props and stop them by either close the fuel levers or shut the fuel valve and then pull the prop lever fully back.

Finn

[Ken Stallings 1]

I´m not sure what You mean Ken.

If You think that the props should stop by only feathering the props You are wrong. The prop will run as long as the engine is running.

I can perfectly feather the props and stop them by either close the fuel levers or shut the fuel valve and then pull the prop lever fully back.

Finn

No, no.

We had this discussion a few weeks ago. Of course you have to shut down the engine. However, go up and try this.

Fly at cruise and then perform the trained EP for an engine that must be shut down. Which is, pull the power to idle, pull the prop condition lever to feather, and pull the mixture to idle cutoff.

In 100% of actual aircraft, that will feather the prop and it will stand tall.

Yet, in the Twin Otter Aerosoft produced, this sequence does not produce a feathered prop. Instead the engine is shut down but the prop windmills.

Now, what you do have with your aircraft is that if I have the autofeather engaged and pull the mixture to idle cutoff then the prop automatically feathers and stands tall. But, that's not correct procedure. And it's not how a pilot recognizes an engine condition warranting emergency shutdown.

Normally, you see an RPM or manifold pressure fluxtuation, or you identify an overheat or onboard fire.

The point is you want to feather and shutdown prior to reaching a low RPM setting. Some props have a set of pins that lock the props in the unfeathered position during engine shutdown. So if the pilot is tardy feathering the prop, the blades will lock when then RPM drop low enough and a feathering is impossible. Then, the hapeless pilot is left with a speed brake (windmilling prop).

Ken

[Ken Stallings 1]

Hi Ken,

My prop rpm is set up on the default keys I think. ctrl-f1 is reduce rpm all the way, ctrl-f4 is advance all the way. ctrl-f2 is reduce slowly, ctrl-f3 is advance slowly.

I was confused at first because when I pressed ctrl-f1 the props just remained at about 75%. BUT when I press ctrl-f2 over and over the props go to feather. This seems quite useful actually, because you can use the ctrl-f1 as a way to get the prop rpm to 75% but not go over into feather.

I'm not sure if this is where you had your problem but hopefully it helps.

cheers,

John

John,

I use a throttle quadrant. So, I don't use keystroked to adjust the blade angles. I just cycle the prop condition levers on my dual engine throttle quadrant (CH Pro). The thing is that when I pull the lever on the controller full back then the virtual prop lever in the DVC is also pulled to the full feathered position.

Cheers,

Ken

[JAL 0]

Hi Ken,

Ok, cool. I didn't realise that was a problem with the original aerosoft file, but I just did the procedure you described using the aerosoft 1.10 file and my own, and my modifications fix that too.

I must say I really appreciate your expertise, it really helps to have your detailed descriptions about how things should be in real life.

Cheers,

John

[JAL 0]

Hi all and especially Ken,

I've just been doing some more testing and trying to do Ken's emergency procedure to feather the props...

Here is the bad news:

In my last post I reported how I was able to feather the props with the new air file but not with the old. Since then I've tried to do Ken's test again by moving the prop levers all the way back in the VC, and guess what, now I can't get it to work properly with either air file. I can't figure it out I must have done something different (wrong) in that last test and I just can't get it to work again.

Here is the good news:

To get a better idea of what Ken is talking about I linked my joystick throttle control to the prop levers to simulate Ken's prop levers. When I do Ken's procedure I see what he means the props do not feather. While in flight it appears that the whole travel of the prop levers equates to 75% to 96% on the graphics in the VC. When I move my "throttle" (i.e. now propeller) lever back and forth the props cycle merrily between 75 and 96 but I never get them to feather when the engine is running or not running. Is this how yours behaves Ken?

But! if I have the prop lever at 75% and then press ctrl-f2 once the props feather right away. Although some might argue that it is not correct that the full travel of the prop lever is only between 75% and 96% I actually think it is more realistic. Here's why:

My Twin Otter manual states in the section about prop levers:

"To prevent inadvertent propeller feathering, a gate stop is incorporated in the quadrant for each lever; when feathering, the appropriate gate stop is overridden by pushing up on the lever and then continuing the movement aft."

Therefore, this safety mechanism has been implemented in the sim by not allowing props to feather with the normal travel of the prop levers. There is no need to worry that your props will inadvertently feather in flight and the torque gauge jump through the roof. By pressing ctrl-f2 (or any other button or key you want to assign) it is like pushing up on the real life prop lever to get it past the gate stop and go into feather.

I hope this resolves this issue, and please try it out for yourself to make sure I've got it right this time!

Cheers,

John