investigating New Landing Gear and Tires Problems

[Art-J]

I'd say It is partially a map issue in a sense that on some maps we're getting "only" exacerbated bouncy-bouncy tailwheel behaviour without structural failures (not a great thing but it's not a functional obstacle), while on the others the tailwheel gets actually damaged.

I don't own Syria map, but if what you say happens on more airbase over there, then it would seem both Ugra-developed maps might be more problematic than ED ones.

[Night Owl]

I only ever got the exaggerated and never-stopping bouncing, not actual damage to the tail wheel. That was on Normandy, Syria and Caucasus

[Slippa]

Seems Night Owl’s seeing the same as I am.

A pretty vigorous perpetual bouncing of the tailwheel. I’m only seeing it in the Mossie. Tried a good few airfields on the Normandy map and get the same thing. Just trying to taxi out I’ve busted the wheel twice. I’m getting the hang of it but it can still swing out depending on the bounce. I’ll try a few more maps, think I tried it from Manston on the Channel and it was the same though.

[Tea-Pig]

I have been finding it difficult to reproduce the tailwheel collapses I had during slow speed taxiing at Villacoublay, but here's a short track of a landing on the Caucuses Landing Practice instant action mission. The landing is I think fairly decent and if the track is stopped before the rollout then no damage is recorded, but the tailwheel fails with it's tyre blow-out sound effect some forty seconds later right down the other end of the runway for no obvious reason I can see. 

Other than this tailwheel fragility I have to say I am rather loving the Mosquito, which along with the A4, and the Huey seem to be particularly well suited to my motion rig.

tail wheel3.trk

[Skewgear]

The tail-down moment that occurs when the main wheels touch down is overdone. Curious to see if anyone can reproduce a smooth touchdown like these sequences of KA114.

 

[grafspee]

1 hour ago, Skewgear said:

The tail-down moment that occurs when the main wheels touch down is overdone. Curious to see if anyone can reproduce a smooth touchdown like these sequences of KA114.

 

What astonish me is that back firing sound, as soon pilot cut throttle this amazing sound comes out of the engines, i wish ED could update that sound, i think they did not upgrade this sound when last time they upgraded ww2 planes sounds. Most of the sound pack is still like +10 years old.

[Terry Dactil]

7 hours ago, Skewgear said:

Curious to see if anyone can reproduce a smooth touchdown like these sequences of KA114.

No you can't. Unfortunately the Mosquito landing dynamics have been completely unrealistic from day one.

Even if you can touch down with a sink rate less than one inch per week you will still get a  strong pitch up moment and bounce off the runway.

IRL with a very small sink rate the drag of the main wheels spinning up will actually produce a small nose down pitch moment that will help stick you on the runway. Otherwise, on touchdown, a small nudge forward on the stick does the job. There is no need for nose down trimming or pitching down to smash onto the runway.

(In case you think this is BS, I have several thousand hours of flying tailwheel aircraft -from Tiger Moths to DC-3s- so I think I can say I know what wheeler landings should be like.)

I complained about  the unrealistic landing dynamics years ago.

 

[stuart666]

Ive landed it without getting a bounce. I grant you it isnt particularly typical that I grease the runway that way...  The only way I found you could do it was use a little throttle to ease rate of decent as you cross threshold, which is a distinctly jet thing to do.

I do think however, and this is gut instinct than any real scientific approach, some maps show worse effects than others. Ive not tried Syria, but its my gut instinct Normandy is bad for this. Falklands however seems to be good.

As for the tailwheel, I was thinking about this the other day. its almost as if the modelling of it (not the 3d model) was eccentric and off centre. So that as you move its the proverbial square wheel. So I think there are a few issues perhaps all being conflated together.

Tail wheel easily damaged. Whcih has always been there, but some of the other elements are making it more apparent.

Eccentric tailwheel modelling. Which is decidedly new, and may be contributing in some environments to easy tail wheel damage.

Excessive bounce. Which is new, and if you have a bit of a pitch up as you land as well as the bounce, then you are setting yourself up to slam the tailwheel on the tarmac with inevitable collapse.

Feel free to add.

Edited January 1 by stuart666

[Yo-Yo]

12 hours ago, Terry Dactil said:

No you can't. Unfortunately the Mosquito landing dynamics have been completely unrealistic from day one.

Even if you can touch down with a sink rate less than one inch per week you will still get a  strong pitch up moment and bounce off the runway.

IRL with a very small sink rate the drag of the main wheels spinning up will actually produce a small nose down pitch moment that will help stick you on the runway. Otherwise, on touchdown, a small nudge forward on the stick does the job. There is no need for nose down trimming or pitching down to smash onto the runway.

(In case you think this is BS, I have several thousand hours of flying tailwheel aircraft -from Tiger Moths to DC-3s- so I think I can say I know what wheeler landings should be like.)

I complained about  the unrealistic landing dynamics years ago.

 

Wheels MOI and inertia effect are taken in account in the model, it's very clear mechanics regarding MOI, inertia, arms, etc. All of them are clearly known.  But another clear mechanics says that there are two factors affecting the nose up-pitch moment (three, to be accurate, but the arm from the landing gear to CoG is well known regarding weight-and-balance docs) - WEIGHT and SINK RATE. 
Modern Mosquito does not land with 9000 kg of gross weight. They are lighter itself (no weapons, ammo, armor, old radio, etc) and they almost never land with full fuel loads.
The second factor is sink rate. Real video shows very low sink rate before touchdown not more than 0.3-0.4 m/s.
DCS planes can be landed on two wheels keeping in mind all these factors.
There is a fourth factor, though, but it is rather human than physical: this small stick movement on touchdown. In RL control inputs based on small accelerations are intuitive and way easier than in simulation, because the control loop acceleration-muscle reaction as a reflex  is much faster than visual (position) - muscle reaction. It is a phenomena of a bicycle - riding a virtual bicycle is close to impossible, even if you have years of RL biking practice.

 

[Terry Dactil]

Thanks Yo-Yo. That's fantastic and exactly what I wanted to see.

Because I can't do that on my system I am now suspecting I have something screwed up in it somewhere as it was installed a long time ago and had many updates.

Therefore I'll be doing a complete clean re-install of the program in the hope of getting what you have there.

Thanks again.

[Skewgear]

Your installation isn't bugged, Terry. I'm unable to reproduce Yo-Yo's video without the kangaroo-hopping antics we're all now familiar with. Aircraft weight set to 17,500lbs (~7,900kg), so in the middle of the 17,000lbs - 18,000lbs range given in the 1950 Pilots' Notes for an approach speed of 125mph. Gun ammo reduced to 0, fuel reduced to below 20%. The approach speed at this weight is the same as in the ED Mosquito module manual.

I've even tried dragging it in for half a mile, skimming the treetops, to minimise vertical speed on touchdown to an absolute minimum. Still it bounces into the air like a rubber ball.

The Mosquito Bug Challenge remains: Can anyone produce a video or track file that shows a smooth DCS Mosquito touchdown? After all, we're assured it's possible...

[Skewgear]

More interesting detail about Mosquito landings. Here is a modified FB.VI, LR359, being landed aboard HMS Indefatigable in March 1944 by the famous test pilot Eric "Winkle" Brown. Note how the aircraft behaves on touchdown - there is little or no sign of a nose-up/tail-down pitch upon touchdown, even with the arrester hook fitted. The shock absorbers also have a noticeably long stroke.

 

LR359 was modified for this exercise. What were the modifications - perhaps extra springy landing gear or non-standard shocks were installed?

Here is Winkle Brown himself talking about the mods to the airframe. He mentions three things, including taking the arrester wire at 78mph during land-based trials of the arrester wire system. He then talks about weight, saying the heaviest British carrier landing hitherto had been the Grumman Avenger at 10,000lbs. He adds: "We were going to start the landings of the Mosquito at 16,000lbs".

Follow what the great man tells us himself about modifications to LR359 for the first ever Mosquito carrier landing:

"... so that was quite a bit there that might go wrong, we just didn't know. Thirdly, there was the question of strength of the aircraft. This is a wooden aeroplane, arrested landings aren't very kind to aircraft structure, and many predicted that either the hook would be torn out or the back end would be torn off. But they only did one bit of strengthening: they put two extra longerons in the fuselage to add a bit of strength and the arrester fit was from a Barracuda-type aircraft. These were the only structural alterations that were made. In fact we had a failure of the arrester hook ..."

The Imperial War Museum adds that LR359 had four-bladed propellers fitted: https://www.iwm.org.uk/collections/item/object/205027293

[Yo-Yo]

2 hours ago, Terry Dactil said:

Thanks Yo-Yo. That's fantastic and exactly what I wanted to see.

Because I can't do that on my system I am now suspecting I have something screwed up in it somewhere as it was installed a long time ago and had many updates.

Therefore I'll be doing a complete clean re-install of the program in the hope of getting what you have there.

Thanks again.

I think, the problem is not in the system or version. For me the main difficulty to land on two wheels is to maintain stable approach with constant and low sink rate at least at last 30 meters of altitude. P-51 has VSI that is easily visible, Mosquito - mmm... sorry. Any other type of flaring requires from me very precise altitude and sink rate estimation, that is a big challenge in virtual world. I do not use VR, so the task becomes even more difficult.

 

[Yo-Yo]

54 minutes ago, Skewgear said:

More interesting detail about Mosquito landings. Here is a modified FB.VI, LR359, being landed aboard HMS Indefatigable in March 1944 by the famous test pilot Eric "Winkle" Brown. Note how the aircraft behaves on touchdown - there is little or no sign of a nose-up/tail-down pitch upon touchdown, even with the arrester hook fitted. The shock absorbers also have a noticeably long stroke.

 

LR359 was modified for this exercise. What were the modifications - perhaps extra springy landing gear or non-standard shocks were installed?

Here is Winkle Brown himself talking about the mods to the airframe. He mentions three things, including taking the arrester wire at 78mph during land-based trials of the arrester wire system. He then talks about weight, saying the heaviest British carrier landing hitherto had been the Grumman Avenger at 10,000lbs. He adds: "We were going to start the landings of the Mosquito at 16,000lbs".

Follow what the great man tells us himself about modifications to LR359 for the first ever Mosquito carrier landing:

"... so that was quite a bit there that might go wrong, we just didn't know. Thirdly, there was the question of strength of the aircraft. This is a wooden aeroplane, arrested landings aren't very kind to aircraft structure, and many predicted that either the hook would be torn out or the back end would be torn off. But they only did one bit of strengthening: they put two extra longerons in the fuselage to add a bit of strength and the arrester fit was from a Barracuda-type aircraft. These were the only structural alterations that were made. In fact we had a failure of the arrester hook ..."

The Imperial War Museum adds that LR359 had four-bladed propellers fitted: https://www.iwm.org.uk/collections/item/object/205027293

I am quite surprised as you compare two-wheels landing and this kind of arrested landing. First of all - the last one was performed as a three point landing to establish arrester wire contact. The second fact is, that the arrester creates a tremendous moment that place the aircraft CoG in line with the hook end and arrester wire hardpoints.

[Yo-Yo]

 

[Bozon]

2 hours ago, Yo-Yo said:

Wheels MOI and inertia effect are taken in account in the model, it's very clear mechanics regarding MOI, inertia, arms, etc. All of them are clearly known.  But another clear mechanics says that there are two factors affecting the nose up-pitch moment (three, to be accurate, but the arm from the landing gear to CoG is well known regarding weight-and-balance docs) - WEIGHT and SINK RATE. 
Modern Mosquito does not land with 9000 kg of gross weight. They are lighter itself (no weapons, ammo, armor, old radio, etc) and they almost never land with full fuel loads.
The second factor is sink rate. Real video shows very low sink rate before touchdown not more than 0.3-0.4 m/s.
DCS planes can be landed on two wheels keeping in mind all these factors.
There is a fourth factor, though, but it is rather human than physical: this small stick movement on touchdown. In RL control inputs based on small accelerations are intuitive and way easier than in simulation, because the control loop acceleration-muscle reaction as a reflex  is much faster than visual (position) - muscle reaction. It is a phenomena of a bicycle - riding a virtual bicycle is close to impossible, even if you have years of RL biking practice.

 

Hi @Yo-Yo,

There are two more factors that affect the 2-wheels touchdown:

1. The amount of sagging of the shock absorbers + the tires themselves.

2. The drag (due to friction) of the tires rolling on the surface. A little extra drag of grass surfaces actually helps.

These help to lift the tail and to some degree counter the rotation moment due to the center of mass.

[Slippa]

I can’t get anything close to landing like some of the clips shown here. Before or since the update so that’s beside the point. I need practice.

Never mind the landing, just taxi out in the Mossie as you normally would. I did and noticed straight away something was up. The tailwheel suspension seems to flutter of its own accord since the update.

I spent some time this morning on The Channel, Marianas and the Normandy maps. I rolled out to taxi, took off and attempted landing from different airfields on all three maps. The bounce is with the Mossie. Some fields effect it more or less but it’s with the aircraft wherever she goes.

What I will say is it’s refreshing to see the Mossie getting some love and attention from people that can do some tinkering. I’m grateful for that at least.

Happy New Year all. 

 

[Yo-Yo]

1 hour ago, Bozon said:

Hi @Yo-Yo,

There are two more factors that affect the 2-wheels touchdown:

1. The amount of sagging of the shock absorbers + the tires themselves.

2. The drag (due to friction) of the tires rolling on the surface. A little extra drag of grass surfaces actually helps.

These help to lift the tail and to some degree counter the rotation moment due to the center of mass.

1. No, because the force, that balloons the aircraft, is not a strut/tire recoil. As the plane CoG moves vertically with V velocity having mV momentum,  its undercarriage have to create impulse equal to this momentum to stop vertical CoG movement DURING the forward stroke. Regardless of sagging, rigidness or softness it will create an impuls of F_average*t, where t is the time of the forward stroke. If there is an arm r between contact point and the CoG, this impuls creates an angular velocity that starts to increase AoA.
So, in angular movement the equation will be F_average*t*r = J*w, and substituting the first equation w = m*V_average*r/J, thus the resulting pitch/AoA angualar velocity  do not depends on t or the stroke length. Increased AoA creates a lot of extra lift and aircraft balloons.

2. Grass has less friction, if you mean the initial moment of wheel acceleration, that helps to decrease AoA. That's why grass is better for taildraggers. If you do not believe me,  imagine a football player perfroming slide tackle on tarmac...
 

[Yo-Yo]

By further investigation the recommendation to have 0.3-0.5 m/s sink rate at touchdown are proven. Though it requires a lot of training without RL cues, FM and tires itself are ok.
Except low sink rate that must be almost constant at least for several seconds. the speed must not be too high or too low - the higher speed is the more extra lift will be due to increased AoA.
My touchdown was at 0.6 m/s and it can be considered as a limit for actual gross weight about 7600 kg to perform a wheeler.

 

P.S. Engines on idle just after touchdown prevent from possible bouncing reducing speed and wing lift due to downwash.

 

[Slippa]

Coincidentally, the yellow line on this graph shows my approximate approach path and trajectory during landing training.

Can anyone else keep it on two wheels for a long roll out?

[Nealius]

Recommended sink rate so small it doesn't even register on the VVI gauge seems weird.

Edited January 3 by Nealius

[Terry Dactil]

7 hours ago, Slippa said:

Can anyone else keep it on two wheels for a long roll out?

Yes. Me.

Since I am arrogant enough to think I know all about how to land a tailwheel aircraft on its main wheels, I reckoned it couldn't be me stuffing up the landing, but must be something wrong with my computer or the program.    Therefore I did a clean install of the release version without any mods. (Big pain having to set up all the controls and key-binds again!)

This made a huge improvement and I now get the same results as the videos above. Still wants to oscillate slightly after touchdown, but the bounce back into the air has gone. The wheels now stay on the runway as they should.

I'll keep track of any changes as I restore my mods to see if they were the problem.

 

*** EDIT ***

I take it all back.

For some reason the brake calibration for the analog axis set the default off value to 50% brakes on. (I had previously reduced the analog brake to prevent nosing over)

Removing this brought back the bounce on touchdown. Taxying seemed OK though without the bouncing tailwheel. All this on the Channel map at Manston.

Bugger!

Edited January 3 by Terry Dactil

[Slippa]

3 hours ago, Terry Dactil said:

Yes. Me.

Since I am arrogant enough to think I know all about how to land a tailwheel aircraft on its main wheels, I reckoned it couldn't be me stuffing up the landing, but must be something wrong with my computer or the program.    

I like the self deprecating honesty and humour, very refreshing. 

Wonder why a clean install changed things though? Feel for ya having to redo inputs and all that, can be a right PITA at times. I’ve got about three copies of my inputs, option settings etc. and I still never get it exactly the same again. I’m due a big cleanup and reinstall myself. Of everything, Windows and all. One of those things I keep putting off as I know I’m looking for trouble.

Save a track next time you land and roll it out, I’m feeling like an amateur compared to the landing in the vid. Much more practice for me..

It still bounces excessively and unnaturally on taxi though.

[Bozon]

On 1/1/2024 at 11:04 PM, Yo-Yo said:

1. No, because the force, that balloons the aircraft, is not a strut/tire recoil. As the plane CoG moves vertically with V velocity having mV momentum,  its undercarriage have to create impulse equal to this momentum to stop vertical CoG movement DURING the forward stroke. Regardless of sagging, rigidness or softness it will create an impuls of F_average*t, where t is the time of the forward stroke. If there is an arm r between contact point and the CoG, this impuls creates an angular velocity that starts to increase AoA.
So, in angular movement the equation will be F_average*t*r = J*w, and substituting the first equation w = m*V_average*r/J, thus the resulting pitch/AoA angualar velocity  do not depends on t or the stroke length. Increased AoA creates a lot of extra lift and aircraft balloons.

2. Grass has less friction, if you mean the initial moment of wheel acceleration, that helps to decrease AoA. That's why grass is better for taildraggers. If you do not believe me,  imagine a football player perfroming slide tackle on tarmac...
 

In #1 the issue is that on the one hand you assume that:

”As the plane CoG moves vertically with V velocity having mV momentum,  its undercarriage have to create impulse equal to this momentum to stop vertical CoG movement”

and on the other hand you say:

“Increased AoA creates a lot of extra lift and aircraft balloons”

which is indeed correct, but this starts acting the moment the wheels touch the ground, and thus the 1st statement that the undercarriage has to produce the impulse to counter mV is inaccurate.

Imagine an undercarriage with a really really long stroke. As the wheels gently touch, the initial force that slows the sink rate and pitches up the plane is small and pitch up is gradual. This increases lift which slows the sink rate AND also creates a pitch down moment (since center of lift is behind center of mass) that reduces the pitch up rate.

In contrast, the case of a very short stroke is closer to your calculation where all the impulse is delivered before angle of attack increases and the undercarriage counters all of mV momentum by itself - this is a “heavy landing” bounce.

In a long stroke the increase of lift while the absorbers are still contracting, plus the pitch down moment that this adds is what soften the bounce and allow the wheeling.

As for #2 I am afraid I failed to deliver my point. I tried to use the term “drag” and not friction, but that was not clear either. What I meant was the force that slows down a plane that is rolling on the ground - this is not strictly the friction between the rubber and the surface but a force caused by the deformation of the tire as it rolls.

An aircraft that is given an initial speed and released to roll along a tarmac runway will go farther than a similar experiment on a grass field or a rough surface with small bumps that deform the tires as they hit them.

This is not the same as the coefficient of friction between rubber and the material of the surface, which only needs to be high enough so the wheels will roll without slipping, and makes little difference beyond that.

[Yo-Yo]

If we want to find a state for a time moment when the plane stops sinking its CoG, the stroke length DOES NOT MATTER. As I wrote before it's only a matter of momentum.  You can expand this process or make it very short - the resulting angular velocity increasing your AoA will be the same. The only difference could be that very long stroke and long time of angular velocity rising allows to react and pull the stick forward.
But anyway, the strokes for any certain plane is determined, so we have to deal with real parameters.

Regarding the second pint: I thought that you meant only initial moment of wheels acceleration at touchdown where rolling friction is not relevant. But, I have to say, that my experiments with the different types of rolling friction for pneumatic gave exactly the same results for quite uneven grass surface (Duxford airfield is much smoother) and smooth  tarmac.
The reason is that in steady movement bumps are averaged.