reported On Vortex Ring State from active Mi-8 instructor

[randomTOTEN]

1 hour ago, cw4ogden said:

but I call your attention to 3 minutes and 30 seconds. This appears to be a high fidelity mi-8 simulator, capable of modeling LTE, so I am making an assumption, that VRS is modeled as well.

I think the more relevant time stamp is actually 3 minutes, 42 seconds. It's immediately before he enters LTE (which I would love to have modeled.) From my viewing, it is the moment where the horizontal IAS is the lowest, and the VSI is simultaneously the highest. I.E. the best candidate for VRS.

 

If I'm reading the instruments correctly, at 3:30 the KmIAS is 100. Vertical speed -4mps. At my time stamp the values are about 60kmIAS and -4mps. Let me know if you think that's a more severe example. I think it is, but I have 0 RL rotor time. My time stamp values translate to 32 KnotsIAS and -787fpm.

1 hour ago, cw4ogden said:

I believe had this been DCS's mi-8 modeling, the pilot would have been eaten alive by VRS.

Here I am in the DCS Mi-8. I have achieved my values. I am not in VRS.

 

This track is 9 minutes long, and taken from the Caucasus Free Flight Instant action. I set internal fuel to full, with no weapons (and I presume no cargo). Please have patience while I refuel, take off, and position above the runway (and get better light on the instruments). Feel free to take control. I vibrate (presumably ETL) but remain in control for extended time.

 

I think this is a much better avenue for discussion than misquoting values to a 3rd party, and thank you for providing concrete values which we can test. In exchange, enjoy my test of the quoted values.

 

Thanks!

Eight VRS Test 60kmh -4mps full fuel .trk

[cw4ogden]

5 minutes ago, randomTOTEN said:

If I'm reading the instruments correctly, at 3:30 the KmIAS is 100. Vertical speed -4mps. At my time stamp the values are about 60kmIAS and -4mps. Let me know if you think that's a more severe example. I think it is, but I have 0 RL rotor time. My time stamp values translate to 32 KnotsIAS and -787fpm.

I meant the timestamp only as a starting point.  We are talking about the same moment though.

[cw4ogden]

15 minutes ago, randomTOTEN said:

I think the more relevant time stamp is actually 3 minutes, 42 seconds. It's immediately before he enters LTE (which I would love to have modeled.) From my viewing, it is the moment where the horizontal IAS is the lowest, and the VSI is simultaneously the highest. I.E. the best candidate for VRS.

 

If I'm reading the instruments correctly, at 3:30 the KmIAS is 100. Vertical speed -4mps. At my time stamp the values are about 60kmIAS and -4mps. Let me know if you think that's a more severe example. I think it is, but I have 0 RL rotor time. My time stamp values translate to 32 KnotsIAS and -787fpm.

Here I am in the DCS Mi-8. I have achieved my values. I am not in VRS.

 

This track is 9 minutes long, and taken from the Caucasus Free Flight Instant action. I set internal fuel to full, with no weapons (and I presume no cargo). Please have patience while I refuel, take off, and position above the runway (and get better light on the instruments). Feel free to take control. I vibrate (presumably ETL) but remain in control for extended time.

 

I think this is a much better avenue for discussion than misquoting values to a 3rd party, and thank you for providing concrete values which we can test. In exchange, enjoy my test of the quoted values.

 

Thanks!

Eight VRS Test 60kmh -4mps full fuel .trk 660.51 kB · 2 downloads

Indeed it looks safe replicating the conditions.  I think his speed dips down below 50 for a brief instant at the end, though.  I need to fly a few test flights based on some things brought up in the discussion, hopefully will generate a better track file.  

[randomTOTEN]

4 minutes ago, cw4ogden said:

I think his speed dips down below 50 for a brief instant at the end, though.

I think mine does too. And I get to nearly -5m/s for a moment there.

 

EDIT: I also judge that he has begun to loose tail rotor control at that point. Something we don't have.

Also, did you notice the airport is Sochi-Adler?

Edited June 8, 2021 by randomTOTEN

[cw4ogden]

@randomTOTEN

Take a look at this track.  It takes three hops to induce, but see if you can find VRS parameters met on that third and final attempt.

VRS2.trk

Edited June 8, 2021 by cw4ogden

[randomTOTEN]

I'll be gone for 2 days, but will try and watch when I get back.

[cw4ogden]

This is the chart for VRS for the mi-8 from a user in the Russian language forums.

I've been hesitant to call this a bug because the possibility could not be ruled out that the mi-8 behaves markedly different than other helicopters for some unknown reason.   With the surfacing of the chart, I think we can rule that out as a possibility.

The points I've made can be seen in this diagram, as well.  Stay below 30 degrees VRS should not be possible.  Stay above 40 km / h, again VRS should not be possible.

I've included two tracks and a second chart, with an additional line drawn to represent 30 degrees and an "X" to indicate where VRS was encountered in the track file 2.trk

The second track file takes three attempts to induce VRS, so the third and final iteration is the one to focus on.



 

 

VRS2.trk 2.trk

Edited June 9, 2021 by cw4ogden

[Hardcard]

I don't know how realistic that simulator is, but sure, that kind of approach will definitely get you killed in DCS  

[cw4ogden]

20 hours ago, Hardcard said:

I don't know how realistic that simulator is, but sure, that kind of approach will definitely get you killed in DCS  

The marked spot on the graph is close to 55 km / h as the lines are graduated in 20 km / h increments.  

The DISS unfortunately loses signal, so it is off for portions of the videos.  It's tough to say where VRS actually starts, but I think you can see in the videos I'm near 60 km / h and about 3 to 4 m/s (on VSI so some lag) when it starts to form, and remain in the 40+ km/h range fully developed.  Additionally, I would expect the transition to a vortex ring state to be accompanied by a loss of nearly all forward speed, and transition to near 90 degree approach angle, i.e. straight down.  VRS susceptibility is a low speed phenomenon, but once you get fully into it, it's a NO SPEED phenomenon.

My assertion is That DCS models VRS something like the annotated diagram posted below, with the curve encompassing airspeeds well beyond 40 km/h and approach angles shallower than would be expected according to the original diagram:

 

 

 

 

 

In the last video, VRS 2, you can see the beginnings of VRS near 70 km/h.  Nearly twice the maximum speed as indicated by the Mi-8 VRS diagram.

Edited June 10, 2021 by cw4ogden

[cw4ogden]

I have found the error, and it appears, it is indeed my fault, and related to my test setup conditions. I intended the test to be done at or near sea level. I was fooled by the cockpit altimeter reading zero in the mi-8, zero set to airfield elevation. These tracks are taken at approximately 6000 feet MSL which would almost certainly explain the deviations from the chart.
 

[randomTOTEN]

On 6/8/2021 at 1:13 PM, cw4ogden said:

see if you can find VRS parameters met on that third and final attempt.

I think you did find VRS paramteters on your final attempt. But you're the helicopter pilot, you can judge these values probably better than I can.

 

You also make analysis hard here. I have no output of raw flight data (no Tacview), and am forced to rely on cockpit instruments. You fly this test like a pilot, and not a test pilot. You are trying to replicate landing attempts that end in VRS. My example shows my attempt to maintain steady flight parameters to compare them to the stated "impossible to perform in DCS" condition. But that simulator flight was also read off instruments.

To counteract the instrument delay I tried to achieve steady state flight (and thus steady state indications) for better judgement.

In your track VRS2, what you perform are probably what you consider normal landings, but they make judgement and measurement difficult. Your speed and vertical descent change continuously. The instrumentation lags, and it is extremely difficult to judge exactly what the conditions of flight are.

You state,

4 hours ago, cw4ogden said:

The DISS unfortunately loses signal, so it is off for portions of the videos.  It's tough to say where VRS actually starts,

DISS does gain signal just before the end of the track, but there is a delay as the needles are driven to the proper positions. Your speed and descent rate are of course continuously changing. You don't know "where VRS actually starts" because you don't control the flight parameters.  They are continuously changing, and it is impossible to carefully judge exactly which parameters are resulting in VRS.

I have to slow down the track greatly just to be able to get analysis of the instrument gauges. In my track I give you several seconds of constant speed and vertical descent at the chosen values. We test a single condition to see if it results in VRS. You throw a wildly variable approach then claim it's broken.

 

People come onto forums like this all the time, claiming this and that is broken. They have a variety of experience. You start this thread by asking your SME about the DCS sim. You give him an incorrect value of VSI, and no doubt he probably took that into consideration into his reply. He's never touched the simulator I understand. We just have to go off your experience. And your tracks. And the graphs, FAA guidance, and other sources.

 

At first I didn't believe you, thought you were just another frustrated pilot crashing. Perhaps you didn't factor the different cues you must use. There is an adjustment that needs to be made for RL pilots to often adapt to "sim world" as the visual, auditory, and stomatographic cues are often missing or lacking. Maybe you don't look at the VSI in RL because you can better judge it outside the cockpit windows. You either have to learn the simulators depiction of the "sight picture" or more closely watch VSI.

 

But I do think you have a case, and it deserves merit. I want to help you with that case.

 

I think better data would help your case greatly, but as I watched this track, three landings isn't enough as I can see. Perhaps the developers have already output the flight data of your tracks and found the problem. I'm just going off what I can see.

 

 

Going to watch your 2.trk now...

[randomTOTEN]

On 6/9/2021 at 10:33 AM, cw4ogden said:

with an additional line drawn to represent 30 degrees and an "X" to indicate where VRS was encountered in the track file 2.trk

I don't agree with this.

I don't believe you were in VRS until less than 35m AGL on the file "2.trk"

I tested this by taking control and attempting recoveries by increasing collective hard. If you were in VRS this would have no effect.

The lowest I can recover by increasing collective is 35m. I judge that you were never in VRS above 35m.

 

2trk recover 35m agl.trk

 

I controlled view and simulation rate earlier. When Radio altimeter indicates 35m I hit "Escape" key on keyboard, and select "Take Control"

 

EDIT: I also think you compound the analysis be introducing ground effect into these tests. Is ground effect negating VRS just as you encounter it? Is the inability to escape the ground below 35m just because of inertia? Or am I actually in VRS?

Edited June 10, 2021 by randomTOTEN

[cw4ogden]

3 hours ago, cw4ogden said:

I have found the error, and it appears, it is indeed my fault, and related to my test setup conditions. I intended the test to be done at or near sea level. I was fooled by the cockpit altimeter reading zero in the mi-8, zero set to airfield elevation. These tracks are taken at approximately 6000 feet MSL which would almost certainly explain the deviations from the chart.
 

@randomTOTEN

I'm not sure if you saw this.  I incorrectly used an airfield at 5500 feet elevation for the testing, thinking I had chosen one near sea level.  I finally noticed it making the videos.  I even stated once or twice it felt like an aircraft operating at 6000.  Turns out I was.

 

[randomTOTEN]

7 minutes ago, cw4ogden said:

I'm not sure if you saw this.  I incorrectly used an airfield at 5500 feet elevation for the testing, thinking I had chosen one near sea level.  I finally noticed it making the videos.  I even stated once or twice it felt like an aircraft operating at 6000.  Turns out I was.

So your entire thread is invalidated because you misjudged the density altitude?

[cw4ogden]

5 minutes ago, randomTOTEN said:

So your entire thread is invalidated because you misjudged the density altitude?

I could not find an explanation for the results I was seeing.  But I was operating under the false assumption I’d setup the test properly.  Knowing the conditions were not sea level, I’d say that accounts for what I was seeing.  
 

I flew several iterations at sea level, after discovering my error.  Had I set the test up properly, I’m not sure I’d have opened this thread, so that’s on me.  
 

it seems to be working as it should and yes, I’d say the density altitude mistake could be enough to account for my most likely invalid findings.  
 

I am willing to test more, but I believe I found the error to be in my testing methods, not the flight model.  

[cw4ogden]

It’s hard to complain that the flight model feels like 6000 feet when your test conditions are actually at 6000 feet

 

everyone has my sincere apologies for all the drama.

[scoobie]

[Deleted. I need to practice reading more ]

 

Regardless of that... I deeply/sincerely believe there is no need to apologize, Sir - you spotted something fishy and dug into it to figure out what the hell was going on. I can't see anything wrong in that - on the contrary! Digging, questioning, challenging, examining - that's the way to go. I can only thank you for that!

 

Edited June 11, 2021 by scoobie

[Extranajero]

15 hours ago, cw4ogden said:

" Something about getting no feedback on VRS "

I had to paraphrase something you said earlier in the thread there. Sorry about that, the threads too long to wade through.

There's a piece of software called Simshaker that reads internal DCS variables and uses them to send low frequency sound files to a Buttkicker or similar. Probably I haven't explained that very well, but what it means in practice is that when the Mi-8 ( or other helos ) start approaching the flight regime where you are going to run into VRS you'll get feedback through the seat of your pants.
It's way more valuable than it sounds because you can have your eyes out of the cockpit - not glued to the VSI - during landings.
 

[cw4ogden]

1 hour ago, Extranajero said:

I had to paraphrase something you said earlier in the thread there. Sorry about that, the threads too long to wade through.

There's a piece of software called Simshaker that reads internal DCS variables and uses them to send low frequency sound files to a Buttkicker or similar. Probably I haven't explained that very well, but what it means in practice is that when the Mi-8 ( or other helos ) start approaching the flight regime where you are going to run into VRS you'll get feedback through the seat of your pants.
It's way more valuable than it sounds because you can have your eyes out of the cockpit - not glued to the VSI - during landings.
 

I've seen a few fellow multiplayer friends order those.  I don't use a home cockpit.  I convert my couch to a cockpit so seat shaker would be an unwelcome addition to the myriad of wired devices and usb hubs that linger at my feet when I'm not playing DCS.

My son is getting a VR for his birthday, and I'm gonna "borrow" it to see if I like VR enough to make the move.  

I appreciate the suggestion though.  But what I was trying to point out was a subtle, nuance of the flight model, that ultimately was attributable to  faulty test setup.  I've long since had difficulty avoiding VRS in the sim.  I only questioned the phenomenon's susceptibility range.  

I don't know if I buy into the haptic feedback argument, which may not be what you are asserting, but I am trying VR soon, so maybe I'll feel differently after.

The reason I don't buy into it is take for example the FOV limitations of the simulator.  Sure a TV screen is a limited field of view, but so do NVGs.  The complete lack of peripheral vision and more specifically the lack of motion parallax cues coming from your periphery translate into a layperson, and trained individuals from being able to spot large speed deviations.  Said another way:  with lack of peripheral vision, it is really hard to detect speed cues, and also to detect drift at a hover.  

But with time, the limitations are overcome by a combination of muscle memory and techniques.  Much like driving a car, or flying a plane or riding a bicycle, flying with a limited field of view of 40 degrees takes a level of practice that is somewhat dependent on training the subconscious.  The analogy here being head tracking most people use.

*It is also one of three very perishable skills, the other two being IFR flight and Navigation.  Most pilots can take years off and hop back in and go day VFR flight essentially.  That aspect is very much analogous to "riding a bike" you really don't forget.  Not the case for NVGs, and pretty strict currency requirements exist, because recently really makes a difference.  But if that's the main or only mode of flight you fly, eventually, you get to that riding the bike phase, and your perception / situation awareness returns to near day flight levels.

Anyhow back to the haptic feedback argument.  Once you overcome the initial adjustment, your sense of the aircraft approximates your sense during day flight.


The example I'll use, and my only mea culpa with respect to this entire thread is:  I asserted early on the flight model felt like it was operating at around 6,000 feet.  Unbeknownst to me, I actually was. 

So my point is, with only a similar style aircraft background, flying from my couch, out of the real world flying game for a decade and with virtually no haptic feedback features, I was able to approximate where the aircraft "felt" like it was flying.  I found the "bug" finally, and was pretty spot with respect to the cause, even if it was my own fault.  

Maybe that's a case of the the sun shines on the dog's butt once in a while, but I'll take it as an affirmation my senses were right, if ultimately, my conclusion was wrong based on my unknown erroneous initial conditions.

And, I can't think of a better testimony to the flight model honestly, to have it unknowingly feel like it's acting at a certain altitude, only to discover that it actually was.

Edited June 11, 2021 by cw4ogden

[Art-J]

I think it might still worth to repeat some of the test at the sea level, Caucasus map (even setting up pressure and temp to ISA conditons in the mission editor), just for the reference. Much valuable info is contained in this thread anyway and the admins marked it as "reported" rather than "correct as is", so some flight model tweaks and improvements might come out of it, especially now when they're finalizing the FM for the Hind.

[DeMonteur]

Well if you think that you encounter the VRS on steep approach with high descend speed you are probably wrong. I also thought that it is VRS than I found out that it is probably just too slow engine governor. 

Helicopter modelled here feels really light comparing to the real one. When I want to descend at 150-120km/h i need to lower the collective pitch under the 3°of pitch setting while IRL you are at 4-5°. Plus when you go under 3° of collective pitch setting engines go idle and it took ages when you need power back and set up like 5° col pitch setting. Thats definitely too slow comparing to RL.

 

Mi-8 have a limitation in manual to do not exceed 4 m/s vertical descend speed when you are flying slower than 40km/h. Entrance to VRS would be at least 6m/s maybe more however, I have never encounter it RL (thank's god)

 

[Snappy]

hey ,

sorry, can somebody just give  me a summary of what the latest developement on this point now? I skimmed over the thread and in the end, there seems to have been some

issues with incorrect testing conditions , but on the other hand the thread has a "reported" tag. So is the VRS behaviour correct now or getting a reduction

sometime in the future?

Thank you very much in advance,

Kind regards

Snappy.

[Ramsay]

On 10/23/2021 at 2:07 PM, Snappy said:

is the VRS behaviour correct now or getting a reduction sometime in the future?

There's been no change, VRS onset is still too abrupt/soon i.e. IRL if you fly a 40° descent path, you'll avoid VRS,

In DCS VRS is more like ...

• airspeed less than "x" m/s
• descent rate more than "-x" m/s

= Enter VRS

In the above diagram, the 30° descent path feels fine in DCS but as the Mi-8 slows down to land, it hits the VRS region, losing all lift (and crashes). The only way to avoid the VRS is to fly a faster/shallower approach.

AFAIK there are no plans to change the current behaviour.

Edited October 24, 2021 by Ramsay
Edit the VRS region in the diagram to better represent how it "feels" in DCS

[Art-J]

@SnappyYou're right about testing conditions and reported tag part. But since noone from development team responded in this, or any other thread, further status remains unknown.

Edited October 23, 2021 by Art-J

[Snappy]

Ok, @Ramsay& @Art-J,

thank you both for your helpful and quick replies! Guess we just have to wait until ED gets around to addressing this, if ever.

kind regards,

 Snappy