43 replies to this topic

[p4g3m4s7r]

Okay, took some measurements.

So, a Jenner with 4 jump jets and a listing of a 20 meter jump is jumping between 1.5 to 2 times its own height in altitude from my measurements on frozen city (it's a lot easier to tell in that map because of all the buildings that give you a really good frame of reference).

The standard numbers I've seen thrown around are that mechs are supposed to be 10 - 20 meters tall, which makes sense if we assume the atlas is 17 meters tall because my jenner can clear that relatively cleanly.

Thus going with the same proportions as standinginfire, I have about a 10x5 m Jenner (probably still a little large).

As standinginfire did, I'll use the drag coefficient of a skier at my mech's top speed of 139 kph (38.6 m/s). This gives a drag force of about 47 kN.

4 jumpjets fire for about 4 seconds (I'm not sure if this changes with number, but w/e) and yield a 20 meter jump height so the total energy input is (31751.5 kg * 9.8 m/s * 20 meters) 6,223 kJ.

When running in a straight line on a level surface (the Caldera in Caustic Valley) a 4 second burn (the maximum) resulted in a deceleration from 139 kph to 98 kph (38.6 m/s - 27.2 m/s)(I think it was somewhere between 96 and 99 but it was pretty difficult to tell). Assuming this deceleration was linear (probably physically wrong but it looks like this is actually what's happening in game) the total distance traveled would be 131.7 meters.

Since I was travelling in a straight line I can compare amount of energy removed from the system by a theoretical drag to the energy left in the "actual" system to determine how much energy the jump jets used to maintain forward movement. Then compare this to that output by the jump-jets and see what the difference in height should be given this energy loss.

So for 131.7 meters and 47 kN to 23 kN the energy absorbed in drag is 4700 kJ.

The system exhibited a loss of 11,908 kJ in kinetic energy (38.6 m/s to 27.7 m/s). (lol, almost forgot to multiply by .5)

So, we can see that something is horribly wrong. It should also be noted that a 5x10m profile is still grossly overestimating the forward facing surface area of the Jenner, given that there's still a lot of empty space there.

I'm not even going to bother figuring out how much height should be lost because at this point it seems like the jump jets are actually actively fighting forward movement of the mech.

Caveats:
I effectively assumed linear deceleration and linear decrease in drag which is wrong, but actually would probably cancel each other out since the drag force would decrease exponentially while the speed would decrease in an inversely exponential fashion.

I also assumed mechs are rated in "tons" as opposed to "tonnes" unlike standinginfire. There's effectively very little difference in the two units, however (http://en.wikipedia.org/wiki/Ton).

If you're wondering why I'm working in energy it's because I no longer have a valid matlab install. Otherwise I might've bothered super geeking out by doing some integration of velocities, accelerations, and forces instead.

This is a video game, and I get that it doesn't matter if the physics are accurate. However, there's clearly absolutely no forward thrust going into a mech when firing jump jets and yet there's an upwards of 40% loss in jump height when you are moving forwards. This is bad both in the sense that Jump Jets are actually more tailored towards mechs that just stand there and use them to pop-tart than they are for light mechs that use them to become more mobile. I disagree fundamentally with this and want to see it changed and that is why I bothered making this post.

Nihtgenga, on 28 April 2013 - 03:15 PM, said:

As long as JJ help in turning, they do NOT only have up thrust; otherwise an airborne mech would never turn - but last time I tried with a 'bucket 7M, it did. Currently, the JJ are relatively close to BT Universe description, they only lack full-free-vectoring, they're limited to up+turning, which seems a bit odd to me (why should they be limited?). Full vectoring on the other hand would allow to use the JJ at ground level to accelerate or break, but then it would e.g. possible to do JJ-strafing, speed-turning, etc. too - which would have other consequences to consider.
Why SHOULD it completely cull JJ-snipers from the game? It should only require more skill to hit (now JJ-sniping is the much better form of longrange fire support compared to LRMs)

I get that, however in an even more ironic twist when you use jump jets to turn while jumping straight up it does not effect jump height.

Also, your point isn't super relevant to the problem of losing momentum when moving forward and jump jetting.

Edited by p4g3m4s7r, 29 April 2013 - 05:57 AM.

[StandingInFire]

Nihtgenga, on 28 April 2013 - 03:15 PM, said:

As long as JJ help in turning, they do NOT only have up thrust; otherwise an airborne mech would never turn

You don't need directional thrust to turn midair, just look at those high divers in Olympics they do lots of maneuvering in the air without thrust, it would just be very ridiculous to animate the mech contorting it self so to change direction mid flight.

EDIT: This is how far I got, mind you I haven't done integrals or polynomials in years, so my brain hurts now:

a(t) = 1/2 * p * v(t)^2*Cd*A/m
D = -1/2 * p * Cd * A / m
Cd =1.05, p = 1.2 kg/m^3 (air at 20 degrees), - since decelerating, A = face area, m = mass
a(t) = D * v(t)^2

v(t) = v(0) + Integral from 0->t of a(t) dt

v(t) = v(0) + Integral from 0->t of D*v(t)^2 dt
v(t) = v(0) + D/3*v(t)^3-D/3*v(0)^3

C = v(0) - D/3 * v(0)^3

The polynomial
-D/3 v(t)^3 + v(t) - C = 0

At this point my brain said stahp.

Edited by StandingInFire, 28 April 2013 - 05:03 PM.

[Helsbane]

Nihtgenga, on 28 April 2013 - 03:15 PM, said:

As long as JJ help in turning, they do NOT only have up thrust; otherwise an airborne mech would never turn - but last time I tried with a 'bucket 7M, it did. Currently, the JJ are relatively close to BT Universe description, they only lack full-free-vectoring, they're limited to up+turning, which seems a bit odd to me (why should they be limited?). Full vectoring on the other hand would allow to use the JJ at ground level to accelerate or break, but then it would e.g. possible to do JJ-strafing, speed-turning, etc. too - which would have other consequences to consider.
Why SHOULD it completely cull JJ-snipers from the game? It should only require more skill to hit (now JJ-sniping is the much better form of longrange fire support compared to LRMs)

Never said it would cull them completely. But it would turn off the 'easy mode' for poptarting, leaving only the good jump snipers to deal with. Even if they added a small amount of forward thrust to JJs, it's still simple to drag a mouse backward, right?

[Tennex]

p4g3m4s7r, on 28 April 2013 - 04:05 PM, said:

4 jumpjets fire for about 4 seconds (I'm not sure if this changes with number, but w/e) and yield a 20 meter jump height so the total energy input is (31751.5 kg * 9.8 m/s * 20 meters) 6,223 kJ.

I'm not even going to bother figuring out how much height should be lost because at this point it seems like the jump jets are actually actively fighting forward movement of the mech.

This is a video game, and I get that it doesn't matter if the physics are accurate. However, there's clearly absolutely no forward thrust going into a mech when firing jump jets and yet there's an upwards of 40% loss in jump height when you are moving forwards. This is bad both in the sense that Jump Jets are actually more tailored towards mechs that just stand there and use them to pop-tart than they are for light mechs that use them to become more mobile. I disagree fundamentally with this and want to see it changed and that is why I bothered making this post.

There are way too many assumptions in the numbers being used in those calculations...

though i don't think MWO uses actual mech tonnage in their physics measurements. (or even that there is a realistic physics engine. i would love some clarification on that from PGI) For example people have said that the tonnage presented is "free tonnage" of a mech. since it doesn't make sense that the weight of a mech consists only of weapons/heatsinks/ammo etc etc. are gyros/armor weightless?

its been a while since i took physics. but if i remember correctly there shouldn't be any reason a mech would jump less high if its moving forward. are there any actual measurements intead of guesstamates about mechs losing height when jumping forward? i'm skeptical about any measurements made by players, since the only way to measure it is by watching the outside form in teh cockpit not very reliable...

im gonna go test it in game

Edited by Tennex, 28 April 2013 - 09:20 PM.

[Tennex]

I did some math for the Highlander.

Mass = 9E4 kg
velocity = 15.55 m/s
Surface area = 162m2 (assuming height of 18m)

Expected drag force = 24700N using drag equation.

In game drag force by finding deceleration during flight:
F=m*a

distance jumped = 70m
Initial velocity = 15.55m/s
Final velocity = 0m/s
time = 5s

kinematic equation:
0=15.55^2 +2a70
acceleration = -0.11m/s2

In game drag force = 9900N

says here the drag generated by the game is less than we would expect from a real life situation. though with these uncertain values (mass, surface area) and possibly inaccurate measurements (distance jumped, flight time).

its very easy to draw any sort of conclusion we want without concrete numbers.

Edited by Tennex, 29 April 2013 - 09:52 AM.

[Disapirro]

Don't know if anyone else noticed this, but if you hit x after liftoff, the upward trajectory seems to get steeper. This allows me to get much higher while running and jumping than if I just hit my jump jets.

[p4g3m4s7r]

Tennex, on 28 April 2013 - 09:15 PM, said:

There are way too many assumptions in the numbers being used in those calculations...

Note my caveats. The assumptions I've made almost exclusively favor an argument against me and yet my results still showed there's a huge amount of missing momentum. Also, the biggest assumption is the forward facing area of the Jenner, again with a super conservative estimate.

I think the next thing to measure would be to find some sort of cliff to run off and verify mech behavior. If I slow down just as much when running off the cliff. If the mech basically comes to a stop in mid air then we can assume that the issue is simply their model of drag. If if slows down just as much when I fire jump jets then jump jets are in fact being used to maintain forward movement.

[p4g3m4s7r]

Tennex, on 28 April 2013 - 10:09 PM, said:

I did some math for the Highlander.
its very easy to draw any sort of conclusion we want without concrete numbers.

Your results confirm my issue with the fact that JJ and their corresponding physics currently favor usage in assault mechs and not light mechs. Also your kinematic equation assumes a constant force. So really there's substantially more force than you're incorporating into your analysis. This is why I used energy analysis.

And apart from all of that, this is a look and feel issue. JJ don't feel like they make light mechs more mobile. They feel super clunky and they allow you to jump over an obstacle if you slam into it first.

[General Taskeen]

Even MW3 had jump physics. Not so much here, on a modern engine even...

[Theodor Kling]

Tennex, on 28 April 2013 - 08:36 AM, said:

also Jump jets don't have a forward vector. just a vertical one

despite how the visuals look

well no. Depending onthe mech theyallow vectored thrust. Most assault jumpers have simple downard facing jets i think, but the spider definetly is build to be close to a LAM.

[cyberFluke]

Without even attempting to follow the brain melting math, I've noticed a difference in jump height dependant on speed when piloting my Cat-C1. The speed is a little more sedate than in a jenner, but something feels "not quite right" with the jump jet math in the game.

Disapirro, on 29 April 2013 - 04:55 AM, said:

Don't know if anyone else noticed this, but if you hit x after liftoff, the upward trajectory seems to get steeper. This allows me to get much higher while running and jumping than if I just hit my jump jets.

Haven't tried this, but I will. If true, this adds weight to the "bug" side of the scale.

Just occurred to me:
If killing all forward motion input (hitting x, letting go of w, depending on your config) as you hit the JJ yields the same decline in forward velocity, but a higher apex of jump, then we do indeed have a bug. thoughts?

[MrVop]

Just tried this on the test server, Hard to draw conclusions. It does FEEL like i'm getting higher apex. but Without a observer its hard to measure.

[FireSlade]

Yeah I think that the JJs are messed up in MWO. They have absolutely no forward vertical thrust and rely on any forward momentum to be imparted by what the game says is your forward speed. If you stand up against a building while at full throttle and jump, once you clear the building you will magically move forward. I agree since the JJs are angled down and to the back, that they should always impart a slight forward momentum. While they’re at it they should also have a slight thrust vectoring like if you press forward (w) you should lose less speed or if you hit backwards (s) then you should slow down faster and get more of a vertical lift (though not quite). Also if you hit a strafe left or right the JJs force you to the left or right slightly like “ \ or / “ instead of only being allow to head in the direction that you were originally heading when you jumped “ | “. It would make it easier to us JJs like their true purpose to cross normally uncrossable areas and to add to maneuverability and make it harder for jump snipers.

[MaddMaxx]

p4g3m4s7r, on 28 April 2013 - 04:05 PM, said:

Okay, took some measurements.

So, a Jenner with 4 jump jets and a listing of a 20 meter jump is jumping between 1.5 to 2 times its own height in altitude from my measurements on frozen city (it's a lot easier to tell in that map because of all the buildings that give you a really good frame of reference).

The standard numbers I've seen thrown around are that mechs are supposed to be 10 - 20 meters tall, which makes sense if we assume the atlas is 17 meters tall because my jenner can clear that relatively cleanly.

Thus going with the same proportions as standinginfire, I have about a 10x5 m Jenner (probably still a little large).

As standinginfire did, I'll use the drag coefficient of a skier at my mech's top speed of 139 kph (38.6 m/s). This gives a drag force of about 47 kN.

4 jumpjets fire for about 4 seconds (I'm not sure if this changes with number, but w/e) and yield a 20 meter jump height so the total energy input is (31751.5 kg * 9.8 m/s * 20 meters) 6,223 kJ.

When running in a straight line on a level surface (the Caldera in Caustic Valley) a 4 second burn (the maximum) resulted in a deceleration from 139 kph to 98 kph (38.6 m/s - 27.2 m/s)(I think it was somewhere between 96 and 99 but it was pretty difficult to tell). Assuming this deceleration was linear (probably physically wrong but it looks like this is actually what's happening in game) the total distance traveled would be 131.7 meters.

Since I was travelling in a straight line I can compare amount of energy removed from the system by a theoretical drag to the energy left in the "actual" system to determine how much energy the jump jets used to maintain forward movement. Then compare this to that output by the jump-jets and see what the difference in height should be given this energy loss.

So for 131.7 meters and 47 kN to 23 kN the energy absorbed in drag is 4700 kJ.

The system exhibited a loss of 11,908 kJ in kinetic energy (38.6 m/s to 27.7 m/s). (lol, almost forgot to multiply by .5)

So, we can see that something is horribly wrong. It should also be noted that a 5x10m profile is still grossly overestimating the forward facing surface area of the Jenner, given that there's still a lot of empty space there.

I'm not even going to bother figuring out how much height should be lost because at this point it seems like the jump jets are actually actively fighting forward movement of the mech.

Caveats:
I effectively assumed linear deceleration and linear decrease in drag which is wrong, but actually would probably cancel each other out since the drag force would decrease exponentially while the speed would decrease in an inversely exponential fashion.

I also assumed mechs are rated in "tons" as opposed to "tonnes" unlike standinginfire. There's effectively very little difference in the two units, however (http://en.wikipedia.org/wiki/Ton).

If you're wondering why I'm working in energy it's because I no longer have a valid matlab install. Otherwise I might've bothered super geeking out by doing some integration of velocities, accelerations, and forces instead.

This is a video game, and I get that it doesn't matter if the physics are accurate. However, there's clearly absolutely no forward thrust going into a mech when firing jump jets and yet there's an upwards of 40% loss in jump height when you are moving forwards. This is bad both in the sense that Jump Jets are actually more tailored towards mechs that just stand there and use them to pop-tart than they are for light mechs that use them to become more mobile. I disagree fundamentally with this and want to see it changed and that is why I bothered making this post.

I get that, however in an even more ironic twist when you use jump jets to turn while jumping straight up it does not effect jump height.

Also, your point isn't super relevant to the problem of losing momentum when moving forward and jump jetting.

You accounted for Caustics 1.2 G gravity field in your calc's right? If your going to do motion/ballistics physics, and have known values, always best to include as many as possible right?

Edited by MaddMaxx, 29 April 2013 - 06:51 AM.

[MrVop]

Is the extra gravity value actually implemented?

[MaddMaxx]

MrVop, on 29 April 2013 - 06:53 AM, said:

Is the extra gravity value actually implemented?

No idea, but apparently some think some things are or doing these Maths by pure guessing, would be bad...

[Argent Usher]

BTW "Guardian ECM, like all features in the game, is very close to where we want it to be." Paul Inouye

GL.

[Disapirro]

MrVop, on 29 April 2013 - 06:23 AM, said:

Just tried this on the test server, Hard to draw conclusions. It does FEEL like i'm getting higher apex. but Without a observer its hard to measure.

Try it with a spider if you have one. It may be that it is more noticable because of the speed and distance you are moving. It also feels that your foward momemtum does not slow down when you do this. It is like your are vectoring the thrust up/down by using w, s, or x after liftoff.

[MaddMaxx]

Argent Usher, on 29 April 2013 - 07:42 AM, said:

BTW "Guardian ECM, like all features in the game, is very close to where we want it to be." Paul Inouye

GL.

Your un-focused bias is showing dude. ECM is what it is so let's just move on.

[Argent Usher]

MaddMaxx, on 29 April 2013 - 08:44 AM, said:

Your un-focused bias is showing dude. ECM is what it is so let's just move on.

Calm down Dude, i play a Jenner (only) since 08.08.12 and if i can still get a JJ stuck bug, memory allocation errors, can see the same balancing issues and their related discussions and have with each patch more non canon features surely i can go frenzy all the time or just say:

"Hey it'z PGI, a developer who starts to develop a MMO game without a network engineer."


Don't expect too much.