Control loading feedback vs. The Real Thing (level D sim)

[Learning2Fly]

With respect to control loading, and forces felt from the control surfaces:

How would you describe the realism when rotating on take-off, to high bank turns, and other instances when the aircraft
requires trimming to relieve control stick pressures?

As for the motion, it seems rather inadequate for the most part. Aside from feeling the landing gear pop down, or the instantaneous
acceleration when throwing power levers forward, do you really benefit from the motion?

There has been talk (rumour?) that motion will no longer be a requirement for recertifying in the future. That kinda makes me sad;
the cool factor of having the jolt on touch-down, or the rumble on roll-out is freakin' cool!

edit: if there is a thread about this, please link me.

[Nark]

How would you describe the realism when rotating on take-off, to high bank turns, and other instances when the aircraft
requires trimming to relieve control stick pressures?

Nothing like the real thing.

The motion is way to sensitive during the roll out. You don't have "seat of your pants" feedback from G forces in the sim.
Control/Trim pressure is none existent in the sim (or very little)

Flying a good profile in the sim, is like almost going home with a stripper.

[Taiser]

You don't have "seat of your pants" feedback from G forces in the sim.
Control/Trim pressure is none existent in the sim (or very little)

It gets better if you use a force feedback joystick. I have the awesome Sidewinder FF stick from Microsoft, unfortunately it's no longer made...not sure what the best FF stick out there is now but it makes the trim forces and g forces feel a lot more realistic, not super realistic but not bad...and when the landing gear locks down, the thump is pretty cool!

As far as motion you need a big screen, the bigger the better, it helps. I've been messing around with x-plane and although it's got a bigger learning curve than M$ Flight simulator I think the planes act a little more realistically!

[justwork]

You don't have "seat of your pants" feedback from G forces in the sim.
Control/Trim pressure is none existent in the sim (or very little)

It gets better if you use a force feedback joystick. I have the awesome Sidewinder FF stick from Microsoft, unfortunately it's no longer made...not sure what the best FF stick out there is now but it makes the trim forces and g forces feel a lot more realistic, not super realistic but not bad...and when the landing gear locks down, the thump is pretty cool!

As far as motion you need a big screen, the bigger the better, it helps. I've been messing around with x-plane and although it's got a bigger learning curve than M$ Flight simulator I think the planes act a little more realistically!

You read he's asking about a Level D simulator correct? A level D simulator is the size of a single detached garage that sits on either electric or hydraulic lifts. The cost is substantially more than a force feedback joy stick and a big monitor...

I've always thought that a level D and the actual plane feel pretty close.

[Mach1]

It is close to but not quite like the actual airplane. There is always that millisecond delay as the command you input goes to the mainframe and back to the simulator and the visuals and hydraulics react.... it's small and yet, you still notice it.

To borrow from the HHGTTG:
After a fairly shaky start to the day, Arthur's mind was beginning to reassemble itself from the shell-shocked fragments the previous day had left him with.
He had found a Nutri-Matic machine which had provided him with a plastic cup filled with a liquid that was almost, but not quite, entirely unlike tea.
The way it functioned was very interesting. When the Drink button was pressed it made an instant but highly detailed examination of the subject's taste buds, a spectroscopic analysis of the subject's metabolism and then sent tiny experimental signals down the neural pathways to the taste centers of the subject's brain to see what was likely to go down well. However, no one knew quite why it did this because it invariably delivered a cupful of liquid that was almost, but not quite, entirely unlike tea.

That sums it up nicely.

[Hawkerflyer]

Flying a good profile in the sim, is like almost going home with a stripper.

[Learning2Fly]

That's right I want to discuss full motion, level D sims. The kind that cost 20 million
not FSX.

Specifically control forces as per the inital post. I'm already aware that the motion could not accurately reproduce anything more than about 1.5 g, and then it would be instantaneous at best.

We do model the sim using accelerometer data, and other parameters from the FDR so I was just curious if the yoke pressure felt realistic. When I practice V1 cuts, it takes both arms, and rudder inputs to maintain control until aileron, and elevator trim are set.

Hopefully not many of you have had to deal with dead engines during flight, but perhaps you can comment on the feedback pressures while
in a turn, pitching up, etc.

Thanks much!

[looproll]

Control/Trim pressure is none existent in the sim (or very little)

I don't know what level D sim you're flying, but when they gave me a trim runaway I could barely hold the controls in place, so I'd say the stick forces are pretty good. I had to ask the non flying pilot to help me as my arms were getting very tired, very quick. You certainly feel feedback if the aircraft is out of trim, much like the real thing. In a little jet, if the trim is not in the right place, the stick forces can get out of hand pretty quick. The sim was very similar to real life in that respect.

[TOGA Party]

Don't worry Taiser, I got your humor.

[Panama Jack]

It is pretty hard to simulate G forces; sure, the Level D's rock around a bit and jolt you around during turbulence, rapid acceleration and deceleration.

I gotta admit that as valuable as Level D's are in my training, I have trouble sometimes convincing myself it is a real airplane; in other words, I sometimes forget things or do thing in a simulator that I just would not do in a real airplane. From conversations with colleagues, it seems that I am not the only one.

[Taiser]

Guess I should have put a at the end!

[Learning2Fly]

If this was a question on your exam, most of you would get a big donut. : )

Maybe it's my fault for a somewhat misleading subject line, so I edited the title for clarity.

Rock on!

[fish4life]

I think in order to get any sort of G forces they would need to make sims like this lexus one

[Learning2Fly]

Wow, that is insane! If they could only incorporate a significant vertical axis, simulating aircraft forces would be more suitable in training. Im not even going to guess the cost of that installation.

There's a video of an indy engine dyno that simulates cornering forces to help the engineers control fluid flow throughout the motor. I will try to find it.

[Learning2Fly]

http://m.youtube.com/watch?v=fv53RbvgfGc

that is the one. A bit off topic, but worth the watch.

Back to control loading!

[True North]

Control forces in a properly maintained and certified simulator should be bang on. The certification process requires that the control forces in the simulator, checked yearly during recertification, match the data provided by the aircraft manufacturer. This is all done via computer modelling but I asked my buds at Westjet and after the computer says it's all good they have one of their experienced guys fly the sim for a subjective evaluation. My personal experience over the years in several types is that it is extremely close, if not exact.

[complexintentions]

It is close to but not quite like the actual airplane. There is always that millisecond delay as the command you input goes to the mainframe and back to the simulator and the visuals and hydraulics react....

Errr...you mean, sorta like the fly-by-wire systems in the actual a/c? I've never noticed this lag you speak of in either machine.

And I have to respectfully disagree, I've always found the Level D's to be very faithful in reproducing the real thing. A few small differences in handling perhaps, but in the end it's intended to teach management of situations, not flight basics. Sure, you practice V1 cuts and engine out approaches but most people will never have the chance to compare those situations to real like (thankfully) so how can people come out and say "oh it's nothing like the real thing"?

The older sims, it's true were not quite realistic.

[Learning2Fly]

Alright, so we're getting some cool discussion happening. Some agree, and some disagree.

Could this be a measure of simulator design, or quality perhaps?

As someone already mentioned, these sims are calibrated fairly closely to actual aircraft data. When I say close, I mean...overlaying
accelerometer graphs of sim. vs. aircraft are withing a hair of each other.

There is always a processing delay between load-cell measurement to hydraulic/electric feedback, however I am not able to detect any
noticeable delay in our sims. It's pretty much pull back, and the sim is pitching; or roll left, and you're feeling the bank angle. There are
accumulators within a few feet of the control rams, so pressure changes are happening very quickly.

I'll likely never - make that: I'll never have the chance to fly the real aircraft, but I've always wondered if I'd be able to land the actual
type if given the chance.

Within a few days I'd like to be able to share some videos of me flying the sim (if the company agrees), just to get some input from all of
you. Hopefully there wont be any disclosure issues by placing video on YouTube for instance. We'll see!

[old_man]

Disregard the soundtrack but.....you know you want to try it.

[Gino Under]

Learning2fly

No Level D simulator in service today needs to be 100% accurate in aircraft handling or fidelity.
It almost goes without saying that one of the reasons a Level D isn't 100% accurate is due to the fact that motion and visual systems come with acceptable latency errors. An example of which may show up when a pilot transitions from IMC to VMC on an approach. Usually the pilot will change from stable flight to pilot induced ocillations due to this latency between visual and motion feedback to the human brain. Stay on instruments to avoid the sloppiness when transitioning to land.

Another reason Level Ds tend to be less than 100% is because there is no g feedback and it's unlikely we'll require g forces in a level D for some time. Despite what the upset recovery training gurus would have us believe.

In fact, many airlines are pushing for "no motion, visual only" as a training cost saving measure. And why not?
While that may seem odd to some, the reasons are due to the fact, beyond an initial type rating, a pilot with at least six months flying in the real aircraft actually doesn't need the motion feedback on recurrent visits to the sim.

Control loading forces are measured by an odd looking contraption hooked up to the Level D flight controls. This device pulls and turns against the motion system to measure results which are compared to a qualification test guide (QTG). These are then validated and if within the manufacturers test data submitted in the level Ds data package, the sim is good to go. Further complementary manual testing known as manual QTGs are gathered and submitted to the authority for completion of the flight control validation and used in control loading of the level D sim. The control loading is as close as it can get to the real McCoy. If it isn't in your sim, snag it. Remember, a sim is a computer. The amount of repositioning during a 4 hour training session can corrupt files and turn the sim into a WTF is going on device.

You asked , "With respect to control loading, and forces felt from the control surfaces:
How would you describe the realism when rotating on take-off, to high bank turns, and other instances when the aircraft requires trimming to relieve control stick pressures?"

Rotation realism in a certified level D matches the QTG for that certified level D device. The simulation of my aircraft is very accurate. Probably not 100% but good enough.
There is no g feedback in steep turns. So, the simulated aircraft by comparison just isn't the same. The simulator accuracy is acceptable to the regulatory authorities and that works for me.

I find the trimming response in a level D can be inaccurate compared to the real aircraft, but again, it's good enough for the folks who let us train in them.

"As for the motion, it seems rather inadequate for the most part. Aside from feeling the landing gear pop down, or the instantaneous acceleration when throwing power levers forward, do you really benefit from the motion?"

I disagree. In most, if not all, the motion IS quite adequate. In fact, the newer simulators with electric motion systems are a tad better in response than the old hydraulic systems. As for the sounds in a level D, they are measued by the manufacturer in a real aircraft in flight and the results are provided to the simulator manufacturer for validation. That includes the gear cycle up and down, the RAT when it is deployed the simulated airflow noise, the simulated depressurization, etc., it's all measured and the decibels of each provided by the OEM. Acceleration and deceleration is also measured. When a level D motion system is accelerated or decelerated it matches the acceleration rates and deceleration rates provided by the OEM. These all serve to validate the motion response of the level D simulator.
If you want to discuss the necessity of motion I'd say there are many instances where you don't need motion. Does a pilot under training really benefit from the motion? I'd say, Yes.

"There has been talk (rumour?) that motion will no longer be a requirement for recertifying in the future. That kinda makes me sad."
As far as I know, motion will be required in the future. You can also look for additional fidelity such as icing modelling and stall behaviors.

As for the visual systems, look for the recent LCOS (liquid crystal operating system) high definition systems. They are un-f**king-real!!

I hope this helps?
Gino Under

[Learning2Fly]

Hi Gino Under (Gino?),

Thanks for that great response. I'd like to elaborate on some of the points you made, and clarify what I meant by some.

An example of which may show up when a pilot transitions from IMC to VMC on an approach. Usually the pilot will change from stable flight to pilot induced oscillations due to this latency between visual and motion feedback to the human brain. Stay on instruments to avoid the sloppiness when transitioning to land.

I've seen these acronyms before, but can't be sure what they mean:

IMC = Instrument Motion Cue, VMC = Visual Motion Cue?

In fact, many airlines are pushing for "no motion, visual only" as a training cost saving measure. And why not?

This is likely what I heard as the “rumour” referenced in my previous reply.

Control loading forces are measured by an odd looking contraption hooked up to the Level D flight controls. This device pulls and turns against the motion system to measure results which are compared to a qualification test guide (QTG).

You would be describing a “Load Cell”, correct?

These are then validated and if within the manufacturers test data submitted in the level Ds data package, the sim is good to go. Further complementary manual testing known as manual QTGs are gathered and submitted to the authority for completion of the flight control validation and used in control loading of the level D sim. The control loading is as close as it can get to the real McCoy.

The certification process is quite lengthy, and detailed around here (as everywhere else I'm sure). There is a huge amount of paperwork, and testing that we undergo every period. The FAA is quite serious about record keeping, and sim performance as you are well aware.
It's scary to think of the consequences if mistakes are made; as severe as fines/and revoking operating licenses. If I'm not mistaken, criminal
charges may also be slapped on depending on the incident. That's pretty nuts!

If it isn't in your sim, snag it. Remember, a sim is a computer. The amount of repositioning during a 4 hour training session can corrupt files and turn the sim into a WTF is going on device.

Ha, I know what you mean. A clean reboot will do wonders for these things. It’s a daily procedure at our facility, and it’s not uncommon to reboot the sim throughout the day due to a hiccup in communication between the CPU racks.

I disagree. In most, if not all, the motion IS quite adequate. ... If you want to discuss the necessity of motion I'd say there are many instances where you don't need motion. Does a pilot under training really benefit from the motion? I'd say, Yes.

I think the point of my statement ran away from me. What I meant to imply was that a level D sim as good as they are, cannot replicate
all the forces necessary to give a first timer a true sense of what a climbing, accelerating turn actually feels like. I didn't want anyone to
think that my comparison was trying to illustrate that a Level D simulator was "just as real" as the aircraft...so that is why the thread title
focused on control-loading.

As for motion in general, I agree that it's beneficial. IE: V1 cuts, or trim run-away. The trainee really gets a sense of urgency when an
engine goes down at 200' AGL as the sim yaws, and banks. As you fight the controls, and apply rudder, and aileron to get back to "wings level",
you have visuals, annunciators, and call-outs stirring up even more chaos. The motion tossing you around, and sliding you sideways in the seat
is icing on the cake for creating the final sensory freak-out to almost make you forget you're floating on a platform. Almost.

Doing that same exercise without motion just wouldn't have the cool factor in my opinion, so yes, I agree with you!

I would love to see the LCOS visual you are describing. We are currently using three generations of visuals, and there's such a difference
between all of them. These developers are doing their homework that's for sure. I'd like to see more 3-D imaging, but I'm certain the processing
speeds, and cooling requirements will be astronomical.

[Gino Under]

Learning2fly

For clarification:
IMC = Instrument Meteorological Conditions, (flight in cloud) VMC = Visual Meteorological Conditions, (visual flight out of cloud)

Control Loading
There are 9 specific static control checks. Essentially these are pitch, roll, rudder, nosewheel steering, rudder pedal steering, pitch trim indicator, pitch trim rate, alignment of cockpit throttle lever position in relation to engine parameters, and brake pedal position. These are all compared to position calibration provided in the data package by the OEM.
Testing of the control loading validates the data and make it as-close-as-it-gets to the real machine. A test and evaluation pilot (usually type rated) further 'subjectively' tests and validates the result. A designated pilot from the airline will usually subjectively 'test' the sim before it is accepted by the airline training facility. I can assure you, even airline pilots get it wrong during subjective testing of simulators.

“The certification process is quite lengthy, and detailed around here (as everywhere else I'm sure). There is a huge amount of paperwork, and testing that we undergo every period. The FAA is quite serious about record keeping, and sim performance as you are well aware.
It's scary to think of the consequences if mistakes are made; as severe as fines/and revoking operating licenses. If I'm not mistaken, criminal charges may also be slapped on depending on the incident. That's pretty nuts!”

If mistakes are made during qualification testing and things end up out of tolerance (when compared to flight test data) then the sim fails to qualify. If it fails to qualify then it simply isn’t put forward to the authority for certification. No fines or criminal charges. So, I’m not sure what your thought is in the previous statement that makes it ‘pretty nuts’.

“I think the point of my statement ran away from me. What I meant to imply was that a level D sim as good as they are, cannot replicate all the forces necessary to give a first timer a true sense of what a climbing, accelerating turn actually feels like. I didn't want anyone to think that my comparison was trying to illustrate that a Level D simulator was "just as real" as the aircraft...so that is why the thread title focused on control-loading.”

Sorry, but again I would emphasize that the control loading in a Level D simulator is a replication of test data collected by (insert Aircraft Manufacturers name here) on a REAL flight test aircraft flown by real test pilots. Except for g loading. There is a limit to g forces in synthetic devices, full stop. They certify the data, provide it to the simulator manufacturer who then creates the device and reproduces the same data for control feedback to the pilot under training. So again, I’m not sure why you think it’s bogus?

For example, Boeing uses a B-737NG model YK131 for its data package. All sim manufacturers use the YK131 data package to create a B737NG full flight simulator for Level D certification.

“As for motion in general, I agree that it's beneficial. IE: V1 cuts, or trim run-away. The trainee really gets a sense of urgency when an engine goes down at 200' AGL as the sim yaws, and banks. As you fight the controls, and apply rudder, and aileron to get back to "wings level", you have visuals, annunciators, and call-outs stirring up even more chaos. The motion tossing you around, and sliding you sideways in the seat is icing on the cake for creating the final sensory freak-out to almost make you forget you're floating on a platform. Almost.”

When I read this paragraph it sounded to me like the pilot you imagine shouldn’t be in the seat in the first place if he/she handles a V1 cut in this manner. Most pilots under training in a Level D handle the scenario you imagine in a very controlled and professional way. The motion cues in this scenario are ‘as-real-as-it-gets’. It certainly gives the pilot confidence in handling procedures should he/she be faced with it on the line. In fact, the FAAs NSP group find it close enough to certify Level D 'zero time' for the pilot doing a type rating in a Level D device.

LCoS visual – Liquid Crystal over Silicon.
They are pretty amazing.
You'll also find DLP systems out there.

I hope this discussion helps and I invite you to continue, even if it’s through an email.
Good luck down the road.

Gino Under

[Learning2Fly]

It's scary to think of the consequences if mistakes are made; as severe as fines/and revoking operating licenses. If I'm not mistaken, criminal charges may also be slapped on depending on the incident. That's pretty nuts!

If mistakes are made during qualification testing and things end up out of tolerance (when compared to flight test data) then the sim fails to qualify. If it fails to qualify then it simply isn’t put forward to the authority for certification. No fines or criminal charges. So, I’m not sure what your thought is in the previous statement that makes it ‘pretty nuts’.

My mistake. I wanted to discuss penalties for techs maintaining the sim, and not the qualification process. I should not have mixed the
two together. Disregard all of that.

There is nothing "nuts" about it; those who are foolish enough to purposely sign-off on faulty test equipment should be dealt with accordingly.

I think the point of my statement ran away from me. What I meant to imply was that a level D sim as good as they are, cannot replicate all the forces necessary to give a first timer a true sense of what a climbing, accelerating turn actually feels like.

Except for g loading. There is a limit to g forces in synthetic devices, full stop. <...>. So again, I’m not sure why you think it’s bogus?

We're on a different page. The above quote is referring to motion, not control loading feedback. I'm in agreement with respect to control loading. We both agree it's close to the real thing.

We also both agree that motion cannot properly simulate high g manoeuvres.

When I read this paragraph it sounded to me like the pilot you imagine shouldn’t be in the seat in the first place if he/she handles a V1 cut in this manner.

Well, that's because I'm describing what I felt during my first V1 cut experience.

It came as a surprise courtesy of a fellow sim tech when testing the simulator...very sneaky of him!

I'm not a pilot; just learning as I go for now which explains the wild ride. Within the next year, I'd like to start the application process
for a PPL. Having some formal flgith training would be excellent right now to practice in the sim before going for lessons.

Thanks for answering the IMC / VMC question as well!

[Mach1]

I'll likely never - make that: I'll never have the chance to fly the real aircraft, but I've always wondered if I'd be able to land the actual type if given the chance.

The short answer to that question is; Yes.

The long answer is; When you start with an airline, you do all your training and flight test in the sim and your first flight in the actual airplane usually has passengers in the back. You have to be very sensitive to notice the differences between the sim and the plane. Then, there are the issues that every airplane out there (the same plane from the same manufacturer) flies slightly different. Small differences in manufacturing, things that have happened after it left the factory, age.... The biggest factors that would effect you are the environmental factors. The weight on every landing can be very different, the winds, the runway conditions, and on it goes. So, you will be able to land the actual aircraft but, it may or may not be pretty.

[Gino Under]

Learning2fly

When you say techs who sign off on a sim, do you mean the morning readiness check? Or, do you mean techs who run electronic QTGs and find the sine waves don't match for a particular test and still sign it off as RFT? If so, it's likely to be caught by either a student OR a qualified SFI. In which case, depending on the issue it could be written up as an A snag and AOG the sim.
Why do you think criminal charges apply? And what do you mean by faulty test equipment? I guess I'm not clear on what you mean. Human error in judgement even subjective judgement is a peripheral reality, even for techs who take care of these devices.

Control loading and motion are interrelated. Force feedback is needed in the controls for proper and proportionate motion response. Sticking to motion, a latency error is acknowledged by RAeS as any ground based FFS with motion capabilities inevitably has severe limitations with respect to motion generation. As a consequence, modifications have to be made to the signals derived from the aeroplane's states to keep the motion system from running into its limits and giving erroneous motion perception to the pilot. Such modifications mean that resulting cues generated through the motion system will inevitably deviate from those experienced in the real aeroplane and so sophisticated use of the available motion capability is required to avoid adverse affects.

The frequency response, leg balance and turn-around bump tests have been the traditional motion tests that have been in existence for many years. Usually a special program is used to run these tests which bypasses the main motion drive software normally used for training purposes. That's because the tests are not checking the types of cues to be sensed on the flight deck (these are done in the functional and subjective test section of the QTG) they are checking the performance of the motion system itself (with its payload).

Gino