Question XR2 Touchdown Vertical Speed

[thepenguin]

Simple Question:

Gear damage depends upon two factors: vertical speed and vessel mass.

What's the formula? I would expect something rather complex, but I hope I will be surprised.

 

[JMW]

Don't know if it helps, cos not for XR2 but I've been using:

if(vspdwn <= -9.5/(GetMass()/15000))
crashdamage = 1;

Where vspdwn = negative vert speed.
-9.5 is max neg vert spd (m/s) before "crash" sounds in Orbiter
15000 is PropellantResource MaxMass

Orbiter doesn't seem to be coded for vessel mass relative to "crash"

 

[Evil_Onyx]

Just aim for less than 1 m/s and you are safe even when fully loaded.

 

[thepenguin]

my knowledge of newtonian physics makes me think a few things:

1. the landing gear damage may be based off of the impulse it must absorb (mass*vertspeed)

2. by a couple of technical calculations (damage to things we know is based off of force, not impulse) I found that another option would be the system based off of kinetic energy as a maximum.

I suppose what's left is to actually test these models, but I need to know what the threshold for damage would be (either in kg*m/s or kg*m^2/s^2, depending on which model is right.)

EDIT: The reason I want to have a predictive algorithm is so that I can land by making the most minimal corrections in course for a given vessel mass. I've crashed way too many times.

 

[Cras]

Just aim for less than 1 m/s and you are safe even when fully loaded.

Or more even. I tend to get to within 3 m/s, dont usually touch down that hard, but I have never had issue.

 

[thepenguin]

Or more even. I tend to get to within 3 m/s, dont usually touch down that hard, but I have never had issue.

I suppose that it is a functional approach, but if I can come down faster, then I can benefit from the time I saved there. Because this is for an autopilot, I can calculate in real time, and only make corrections if necessary.

 

[Cras]

I suppose that it is a functional approach, but if I can come down faster, then I can benefit from the time I saved there. Because this is for an autopilot, I can calculate in real time, and only make corrections if necessary.

I always de-orbit the XR2 dead stick so I do the HAC and steep glideslope , the whole thing, and on the preflare that is my target descent rate when crossing the threshold, hasnt failed me yet. Flying a powered landing you certainly can do much much better, if not grease it everytime. Its pretty easy to grease a dead-stick landing too but it really isnt my intention, I just want to make sure I get the thing on the ground without the gear buckling.

 

[dgatsoulis]

Simple Question:

Gear damage depends upon two factors: vertical speed and vessel mass.

What's the formula? I would expect something rather complex, but I hope I will be surprised.

You were right in your assumption, vertical speed and total mass.
From the XR Flight Operations Manual:

Fully Loaded Mass w/14-day LOX and full crew complement (no cargo): 35,748 kg
Maximum Payload Mass: 10,795 kg (max recommended)
Max touchdown descent rate, max load (full fuel, full cargo): 2.8 m/s

[math]\mathrm 2.8\cdot (35748+10795) = 130320.4 \;kg\cdot m\cdot s^{-1}[/math]

The 35748 kg FLM includes 16080kg (empty mass) + 14200kg (Main+RCS) + 3350kg (SCRAM) + 1024kg (14 crew) + 268kg (APU) + 364kg (LOX) + 462kg (???) {I have no idea where those extra 462 kg are from. A fully loaded default XR2 with no cargo, actually weighs 35286 kg}

Max touchdown descent rate, typical load (33% fuel, 75% cargo):4.2 m/s

[math]\mathrm 4.2\cdot (16080+(14200*0.33)+1024+268+364+462+(10795*0.75)) = 130117.05 \;kg\cdot m\cdot s^{-1}[/math]
{Here I am making the assumption that all the SCRAM fuel has been used}

For an autopilot, I'd go for [math]\mathrm M_{total}\cdot (-vel_y) < 13e4 \;kg\cdot m\cdot s^{-1}[/math]

Notice that an empty XR2 with only the pilot, 1 kg of LOX and 5kg of APU can land with a vertical speed:

[math] \mathrm -vel_y = \frac{13e4}{16080+78+1+5} = 8.04\; m\cdot s^{-1}[/math]

IIRC that value is correct. I have landed empty Ravenstars with vertical speeds of about 7.5 m/s without the gear collapsing.

 

[thepenguin]

You were right in your assumption, vertical speed and total mass.
From the XR Flight Operations Manual:



[math]\mathrm 2.8\cdot (35748+10795) = 130320.4 \;kg\cdot m\cdot s^{-1}[/math]

The 35748 kg FLM includes 16080kg (empty mass) + 14200kg (Main+RCS) + 3350kg (SCRAM) + 1024kg (14 crew) + 268kg (APU) + 364kg (LOX) + 462kg (???) {I have no idea where those extra 462 kg are from. A fully loaded default XR2 with no cargo, actually weighs 35286 kg}



[math]\mathrm 4.2\cdot (16080+(14200*0.33)+1024+268+364+462+(10795*0.75)) = 130117.05 \;kg\cdot m\cdot s^{-1}[/math]
{Here I am making the assumption that all the SCRAM fuel has been used}

For an autopilot, I'd go for [math]\mathrm M_{total}\cdot (-vel_y) < 13e4 \;kg\cdot m\cdot s^{-1}[/math]

Notice that an empty XR2 with only the pilot, 1 kg of LOX and 5kg of APU can land with a vertical speed:

[math] \mathrm -vel_y = \frac{13e4}{16080+78+1+5} = 8.04\; m\cdot s^{-1}[/math]

IIRC that value is correct. I have landed empty Ravenstars with vertical speeds of about 7.5 m/s without the gear collapsing.

Wow. looks like I was right about the maximum impulse on landing theory.

mass*vertspeed < 130117.05 should do the trick here.

 

[RisingFury]

I suppose that it is a functional approach, but if I can come down faster, then I can benefit from the time I saved there. Because this is for an autopilot, I can calculate in real time, and only make corrections if necessary.

You can come down faster. Just make sure to pull up before touchdown. I did my OBSP autopilot like this:

First the vessel is set to descent at 5° angle at 150 m/s. These two values are currently hard coded, but will become vessel specific later on, before OBSP is released. The vessel aims a few hundred meters short of the runway and a hundred or so above. When it reaches that point, it starts aiming for the end of the runway, but kills thrust. The velocity drops as the autopilot tries to aim the velocity vector at the end of the runway.

A few moments after touchdown, wheel brakes are applied and are released when the vessel falls below 10 m/s or 20 m/s - I don't remember the exact values right now. The thinking is that when the vessel gets to a manageable velocity, the taxi autopilot takes over and guides the vessel to its parking spot.

I'm currently working on a bunch of work that needs to be done to support ATC. The ATC is currently somewhat dumb and it always sends the vessel along the shortest taxiways it finds - assuming any are defined. Neither the vessel nor the ATC check for any potential collisions that might take place. But that's why ATC is sorely needed before the first release of OBSP...

 

[thepenguin]

I'm currently working on a bunch of work that needs to be done to support ATC. The ATC is currently somewhat dumb and it always sends the vessel along the shortest taxiways it finds - assuming any are defined. Neither the vessel nor the ATC check for any potential collisions that might take place. But that's why ATC is sorely needed before the first release of OBSP...

That sounds really cool! is there any way for me to get involved with OBSP? It sounds like if it were combined with OMP, then we could have multiplayer combat in orbiter.

I'm currently working on a rewrite of the default aap script to be able to do a KSC->WIN flight (and be able to fly out of the atmosphere). I'm currently rewriting the parts of it that handle altitude control.