Simulators & sustained lateral acceleration

Breaking news, useful data or technical highlights or vehicles that are not meant to race. You can post commercial vehicle news or developments here.
Please post topics on racing variants in "other racing categories".
User avatar
Wazari
623
Joined: 17 Jun 2015, 15:49

Simulators & sustained lateral acceleration

Post

Cold Fussion wrote:Why is he wearing a helmet in a simulator?
Time in simulator is also used to condition drivers physically and mentally. Wearing of helmet is almost always done on long runs in simulator to mimic neck fatigue of actual track time.
“If Honda does not race, there is no Honda.”

“Success represents the 1% of your work which results from the 99% that is called failure.”

-- Honda Soichiro

User avatar
Tim.Wright
330
Joined: 13 Feb 2009, 06:29

Re: 2016 Mclaren F1 Team - Honda

Post

I don't think it would be very useful for that. Simulators can't acheive the sustained lateral accelerations which cause the neck fatigue.

If it is a dyamic simulator I'd agree that a helmet would be a good idea for safety reasons.
Not the engineer at Force India

User avatar
Wazari
623
Joined: 17 Jun 2015, 15:49

Re: 2016 Mclaren F1 Team - Honda

Post

Tim.Wright wrote:I don't think it would be very useful for that. Simulators can't acheive the sustained lateral accelerations which cause the neck fatigue.

If it is a dyamic simulator I'd agree that a helmet would be a good idea for safety reasons.
Although this is one of Ferrari's simulators, they are dynamic. Honda has a similar one. Granted I'm nowhere near the shape these drivers are in, my neck was sore for a month after "driving" 4 laps in one. I don't know how many "g's" they can simulate but they have tremendous accelerations in all directions.

http://www.f1talks.pl/2011/12/27/w-serc ... r-ferrari/

BTW, Honda spent over $4,500,000 for one simulator over 3 years ago.
“If Honda does not race, there is no Honda.”

“Success represents the 1% of your work which results from the 99% that is called failure.”

-- Honda Soichiro

User avatar
Tim.Wright
330
Joined: 13 Feb 2009, 06:29

Re: 2016 Mclaren F1 Team - Honda

Post

Dynamic simulators necessarily have to scale the accelerations down because they need to fit inside a room. If you want all the movements that the driver feels to be 100% accurate, you need to make the simulator room as big as the track so you can copy the movements 1:1.

Consider that a single corner at 3G for 3 seconds will need over 100m of lateral space to create this acceleration. If you want accurate longitudinal acceleration then you can add kilometers to this. There are very few simulators in existence big enough to do that. And the ones that are big enough to do it usually don't have the required power to maintain that level of acceleration. For example Toyota's simulator which is good for about 1G if I remember correctly:
Image

If they are only using a hexapod I guess their max accelerations would be less than 1G. If they have an XY table as well (like the Toyota one) they might get up above 1G but only until they run out of space. Enough to make our necks sore but I doubt their drivers would feel anything. They use specific machines for that type of conditioning:
Image

Tilting the cockpit to simulate accelerations is also generally not an option for professional drivers either because they feel the pitch and roll movements and it stuffs up all their feedback.

What's more important in driving simulators is the phasing between the lateral and the yaw movement. This can (potentially) be captured even with lateral accerations and yawrates much less than reality.
Not the engineer at Force India

trinidefender
317
Joined: 19 Apr 2013, 20:37

Re: 2016 Mclaren F1 Team - Honda

Post

Tim.Wright wrote:Dynamic simulators necessarily have to scale the accelerations down because they need to fit inside a room. If you want all the movements that the driver feels to be 100% accurate, you need to make the simulator room as big as the track so you can copy the movements 1:1.

Consider that a single corner at 3G for 3 seconds will need over 100m of lateral space to create this acceleration. If you want accurate longitudinal acceleration then you can add kilometers to this. There are very few simulators in existence big enough to do that. And the ones that are big enough to do it usually don't have the required power to maintain that level of acceleration. For example Toyota's simulator which is good for about 1G if I remember correctly:
http://www.zercustoms.com/news/images/T ... ator-5.jpg

If they are only using a hexapod I guess their max accelerations would be less than 1G. If they have an XY table as well (like the Toyota one) they might get up above 1G but only until they run out of space. Enough to make our necks sore but I doubt their drivers would feel anything. They use specific machines for that type of conditioning:
http://abload.de/img/f11jrx4s.jpg

Tilting the cockpit to simulate accelerations is also generally not an option for professional drivers either because they feel the pitch and roll movements and it stuffs up all their feedback.

What's more important in driving simulators is the phasing between the lateral and the yaw movement. This can (potentially) be captured even with lateral accerations and yawrates much less than reality.
Quick note about tilting simulators. Ever year we have to go and do recurrent flight training/testing in level D aircraft specific simulators. I attached a video as just an example to show what a level D sim looks like and how it moves.

As you can see the simulator does not move back and forth, it only tilts and moves up and down.

The tilting motion can be used to simulate some form of lateral acceleration or it can be used to simulate the aircraft pitching up and down, it all depends on what the human body inside sees on the screen. Your brain takes many cues when trying to balance the body and if your eyes are open tends to rely on your eyes the most. Therefore when you are in the simulator such as that shown in the video and have no external references a tilting motion can be perceived totally as positive and negative G.

For example if we do a rolling touchdown and apply the brakes in the sim the whole sim will tilt forward. From the outside this doesn't seem like it makes sense as you would think your brain would realise you are pitching down and not actually "slowing down" and feeling the negative G's from that. However from inside the sim the pilot is presented with a 180 degree wrap around screen and that is, to their brain at least, perceived as an external reference. So therefore your eyes see yourself slowing down on a realistic screen. You feel yourself being restrained by the seat belts (coming from the tilting forward) and the screen is showing you that you are on level ground. Your brain perceives this to actually tell you that you are in fact in a moving aircraft and are actually slowing down.

So yes, while a tilting simulator cannot simulate the sustained G forces, it can be very good for simulating how a car feels in minute details such as how it breaks grip, going over bumps, stability under braking etc.

P.s. I am making no claim that a tilting simulator can make any force even close to a race car it can still be significant (obviously less than one G).


User avatar
Tim.Wright
330
Joined: 13 Feb 2009, 06:29

Re: 2016 Mclaren F1 Team - Honda

Post

trinidefender wrote:
The tilting motion can be used to simulate some form of lateral acceleration or it can be used to simulate the aircraft pitching up and down, it all depends on what the human body inside sees on the screen. Your brain takes many cues when trying to balance the body and if your eyes are open tends to rely on your eyes the most.
This doesn't work so well in a car simulator because the lateral and longitudinal accelerations appear very quickly. If you tilt the cockpit at this speed the driver feels the tilt too much no matter what the screen shows. This is due to the accelerometers and gyros that you have in your ear which are sensitive to rotational velocities. In professional sims they usually cut out the tilting effect for this reason. Its more important that the driver feels the pitch and roll correctly.
Not the engineer at Force India

trinidefender
317
Joined: 19 Apr 2013, 20:37

Re: 2016 Mclaren F1 Team - Honda

Post

Tim.Wright wrote:
trinidefender wrote:
The tilting motion can be used to simulate some form of lateral acceleration or it can be used to simulate the aircraft pitching up and down, it all depends on what the human body inside sees on the screen. Your brain takes many cues when trying to balance the body and if your eyes are open tends to rely on your eyes the most.
This doesn't work so well in a car simulator because the lateral and longitudinal accelerations appear very quickly. If you tilt the cockpit at this speed the driver feels the tilt too much no matter what the screen shows. This is due to the accelerometers and gyros that you have in your ear which are sensitive to rotational velocities. In professional sims they usually cut out the tilting effect for this reason. Its more important that the driver feels the pitch and roll correctly.
I know how the human body uses the inner ear to balance itself and I can guarantee you that if it is done properly your brain will totally reject anything your inner ear is telling you. In fact your inner ear partially relies on fluid movement to provide a sense of balance, this can be easily tricked especially by positive and negative g forces.

When your brain is faced with 2 conflicting sources of information between the inner ear and your eyesight it will pick what your eyes are telling you 99 times out of 100.

In those sims you have the cockpit and the screens in front of the driver. From all the sims I've seen (except the picture of that Toyota sim) the screen moves independent of the cockpit. In aviation sims the whole room or pod (for want of a better word) moves. Inside the pod there will be the cockpit and then the wrap around screen and no other visual cues for your brain to use.

Every time I go in the simulator (which I have done quite a few times) it becomes impossible for you to tell that the sim is using tilting to generate loads in different directions. The human brain is simply wired to trust your eyesight more than inner ear, it is actually one reason why so many GA (general aviation) pilots without instrument ratings (a rating that allows you to fly on nothing but your instrumentation) end up crashing when hey go through clouds or in bad visibility, they simply fly straight into the ground and never realise until it is to late. That is how much your brain is wired to trust your eyesight even over your inner ear.

Just as an example, when we are hovering in the sim we accurately feel the movement in the sim to simulate the tiny movements that we input in the controls. We can also feel things like turbulence and when low it correlates to what you see in the screen. That is how accurate and quick reacting these tilting sims are.

User avatar
Tim.Wright
330
Joined: 13 Feb 2009, 06:29

Re: 2016 Mclaren F1 Team - Honda

Post

trinidefender wrote:When your brain is faced with 2 conflicting sources of information between the inner ear and your eyesight it will pick what your eyes are telling you 99 times out of 100.
I wouldn't agree with that at all. Like I said before, It works for aircraft simulators because of the super low frequency and amplitude of the lateral and longitudinal accelerations mean that you can't feel the pitch and roll velocity. I've had the experience of a session with a driver in a dynamic simulator and the first thing he felt was the excessive roll velocity coming from this effect - the solution was to turn it off. In addition to that all of the major simulator manufacturers that I've spoken to say that this effect needs to be turned off for professional drivers because they feel the rolling more than the lateral acceleration.

You can put numbers on it too. If you consider the chicane Roggia in Monza - it's basically a sinusoidal steering input at about 0.5Hz. So in 2 seconds there is a positive and negative peak of lateral acceleration which could be at around 3.5g. Obviously you can't simulate 3.5g by tilting. You can "approximate" around 1g by tilting to 45deg so if you try that at 0.5Hz you will get a peak roll velocity of 140deg/s which is around 18 times higher than the roll velocity of the vehicle itself (which is around 8deg/sec assuming a roll gradient of 0.7deg/g). You can't expect that amount of movement to be filtered out by the visual information.

Other than that, adding additional roll and pitch movements which are in conflict with what the pilot sees is exactly what makes them throw up in your simulator...
Not the engineer at Force India

trinidefender
317
Joined: 19 Apr 2013, 20:37

Re: 2016 Mclaren F1 Team - Honda

Post

Tim.Wright wrote:
trinidefender wrote:When your brain is faced with 2 conflicting sources of information between the inner ear and your eyesight it will pick what your eyes are telling you 99 times out of 100.
I wouldn't agree with that at all. Like I said before, It works for aircraft simulators because of the super low frequency and amplitude of the lateral and longitudinal accelerations mean that you can't feel the pitch and roll velocity. I've had the experience of a session with a driver in a dynamic simulator and the first thing he felt was the excessive roll velocity coming from this effect - the solution was to turn it off. In addition to that all of the major simulator manufacturers that I've spoken to say that this effect needs to be turned off for professional drivers because they feel the rolling more than the lateral acceleration.

You can put numbers on it too. If you consider the chicane Roggia in Monza - it's basically a sinusoidal steering input at about 0.5Hz. So in 2 seconds there is a positive and negative peak of lateral acceleration which could be at around 3.5g. Obviously you can't simulate 3.5g by tilting. You can "approximate" around 1g by tilting to 45deg so if you try that at 0.5Hz you will get a peak roll velocity of 140deg/s which is around 18 times higher than the roll velocity of the vehicle itself (which is around 8deg/sec assuming a roll gradient of 0.7deg/g). You can't expect that amount of movement to be filtered out by the visual information.

Other than that, adding additional roll and pitch movements which are in conflict with what the pilot sees is exactly what makes them throw up in your simulator...
You could very well be right however many of the control inputs done hovering a helicopter or doing long line operations are quite quick, much higher than that achieved in a fixed wing simulator.

Do you have a representative picture of what these simulators look like? Is there any outside references available to the driver at all other than the screen and the cockpit itself?

I am coming purely from an aviation point of view. Oddly enough I've never felt sick in the simulator. However there are those I know who are fine flying yet put them in an older simulator and no matter what they feel sick, I am told that they feel sick in those sims because there is a very slight delay in what your eyes see and what your body feels. The newer sims don't seem to cause people problems, not people who don't feel sick in the actual aircraft anyway.

Tommy Cookers
616
Joined: 17 Feb 2012, 16:55

Re: Simulators & sustained lateral acceleration

Post

human so-called balance organs are a triaxial accelerometer array seeking references variously but only holding them for 4 seconds
this 4 second reference life is the cause of all the problems
and accounts for the difference in motion simulation between motor sport and aviation
in a manoeuvre no longer than 4 seconds the sense of balance is reliable

mandatory 'instrument appreciation' training for fixed-wing pilots dramatically demonstrates this sense-of-balance failure after 5 seconds
UK people will remember cases of motor sport stars who died flying helicopters in conditions of poor or misleading visual references

ie we basically sense changes in acceleration rather than acceleration as such
aviation simulators use this (ie natural rereferencing) to make slow repositioning (after 4 seconds) imperceptible, and so don't run out of travel
Last edited by Tommy Cookers on 19 Feb 2016, 19:50, edited 2 times in total.

User avatar
Andres125sx
166
Joined: 13 Aug 2013, 10:15
Location: Madrid, Spain

Re: Simulators & sustained lateral acceleration

Post

Complete ignorant of simlators here, but when reading this interesting discussion and the problem of feeling a fast roll or not, I´ve noticed there´s an important difference between a car simulator and a plane or helicopter simulator wich may be relevant

Cars move sideways without attitude changes (compared to aircrafts) and those side movements are instant once the driver moves the wheel. But on planes and helis, when you want to move sideways (flat turns apart), you must first roll the aircraft, then comes the movement, so there´s some time for the simulator to roll. This must be way easier to replicate than the lateral movement of a car wich is instant and forces the simulator to do much faster roll movements

Specially on helicopters there´s a delay between the roll movement and the real lateral movement of the aircraft. Planes are in between, not instantly, but neither with a clear delay, just the roll, then the lateral movement. Also, planes and helis do not suffer big lateral Gs, as the aircraft is tilted for turns and most Gs are felt as positive, while cars as we all know generate big lateral forces, not only instantly, but much biggers

User avatar
Tim.Wright
330
Joined: 13 Feb 2009, 06:29

Re: Simulators & sustained lateral acceleration

Post

Cars also respond laterally with a delay of lateral acceleration, yaw rate and roll rate from the input of steering angle. These delays are typically in the order of a few tenths of a second and are responsible for a large amount of the handling "feeling". For this reason you need to be really careful when you are reproducing these movements because even if you don't get the magnitudes right, the phasing needs to be quite good otherwise the driver feels the wrong slip angle.
Not the engineer at Force India

J.A.W.
109
Joined: 01 Sep 2014, 05:10
Location: Altair IV.

Re: Simulators & sustained lateral acceleration

Post

How many lateral G's do you need your simulator to make?

See: https://www.youtube.com/watch?v=xpXxwJNaZDY

NASA might be up for it..
"Well, we knocked the bastard off!"

Ed Hilary on being 1st to top Mt Everest,
(& 1st to do a surface traverse across Antarctica,
in good Kiwi style - riding a Massey Ferguson farm
tractor - with a few extemporised mod's to hack the task).

User avatar
Andres125sx
166
Joined: 13 Aug 2013, 10:15
Location: Madrid, Spain

Re: Simulators & sustained lateral acceleration

Post

Tim.Wright wrote:Cars also respond laterally with a delay of lateral acceleration, yaw rate and roll rate from the input of steering angle. These delays are typically in the order of a few tenths of a second and are responsible for a large amount of the handling "feeling". For this reason you need to be really careful when you are reproducing these movements because even if you don't get the magnitudes right, the phasing needs to be quite good otherwise the driver feels the wrong slip angle.
I should have worded it better, everything has a delay obviously, what I meant is for a car you only have some tenths to roll the simulator, while on a heli that delay might be over a full second, so there´s a lot more time for the simulator to roll

Basically what I meant is the fast roll movement you were talking about wich may be noticeable for some drivers might be a problem for a car simulator (fast rolls for lateral Gs), but not for a heli simulator that have more time to do a controlled and slow roll

J.A.W.
109
Joined: 01 Sep 2014, 05:10
Location: Altair IV.

Re: Simulators & sustained lateral acceleration

Post

I can feel a difference - in the G-forces imposed by being flung around by heavy surf at the beach,
or hard cornering on a motorcycle , when it feels natural to keep binocular vision inline with the angle of lean,
& to compare it with the 'drift' sensation of lateral G-force in a car when tyre adhesion is deliberately overcome..

The steering inputs required seem to come naturally, if you are 'in the groove' or may be awkward/difficult - if not..

It would take a serious simulator to incorporate these actually felt forces .

Funnily enough, the inner ear training that results - seems to help with balance/orientation when walking in the dark,
doing martial arts practice - or even when eyes are held shut due to shampoo in the shower - just as well..
"Well, we knocked the bastard off!"

Ed Hilary on being 1st to top Mt Everest,
(& 1st to do a surface traverse across Antarctica,
in good Kiwi style - riding a Massey Ferguson farm
tractor - with a few extemporised mod's to hack the task).