KSP Challenge

[sal_vager]

Let's stay on topic shall we

[John FX]

5 hours ago, sal_vager said:

Let's stay on topic shall we

I have to say this is how all the challenges I have set have gone, people argue about rules and other stuff so much I end up closing them. As I don't like to post in a challenge without entering I will now make my attempt.

I thought it would be a bit more challenging to do it from 350km up...

[Red Shirt]

My attempt...

.....So close.

[Pand5461]

Nailed it!
Starting in an equatorial orbit. Because having a mothership on a polar orbit is not very practical.

Spoiler

 

 

Transfer:

Spoiler

 

 

Approach:

Spoiler

In the hole:

Spoiler

And a smooth touchdown:

Spoiler

And kOS code to handle this, if anyone cares:

Spoiler

local g0 to Kerbin:mu / Kerbin:radius^2.
print "Standard gravity acceleration: " + round(g0,3).
local Ueff to 154 * g0 * cos(5).
local m0 to 319.5.
local mend to m0-162.
local dob to 240 / 72000 * Ueff.

local dVtot to Ueff * ln( m0/mend ).

wait until body:name = "Moho".

wait 1.

local Vnow to velocity:orbit:mag.

local dVrem to sqrt( dVtot^2 - Vnow^2 ).

print "Available deltaV: " + round(dVtot) + "  dV for polar orbit: " + round(dVrem).

local rho to altitude + body:radius.
local R0 to body:radius + 1000.

local enew to ( dVtot^2 - Vnow^2 ) / 2 - body:mu / rho.

local anew to -body:mu / 2 / enew.
print "New SMA: " + round(anew/1000).

if 4*anew < R0 + rho + sqrt(R0^2 + rho^2) { print "ERROR: NEED MOAR DELTAV". print 1/0. }

local alpha to ((2*anew - rho)/1000)^2 - (rho/1000)^2.
local beta to ((2*anew - R0)/1000)^2 - (R0/1000)^2.

local qa to 1 + (R0/rho)^2.
local qb to R0/1000 * ( (beta - alpha)/(rho/1000)^2 - 2 ).
local qc to ( (beta - alpha) / 2 / (rho/1000) )^2 - beta.

local det to qb^2 - 4*qa*qc.

local ytr to ( sqrt(det) - qb ) / 2 / qa.

local xtr to ytr*R0/rho + (beta - alpha) / 2 / (rho/1000).

local ff to sqrt(xtr^2 + ytr^2).
local rangl to arctan2(ytr,xtr).
local ecc to ff * 500 / anew.

set xtr to xtr*1000.
set ytr to ytr*1000.

print "New axis XY: " + round(xtr/1000) + "  " + round(ytr/1000).

local bnew to anew * sqrt(1 - ecc^2).
local fp to anew * (1 - ecc^2).
local ffvec to V(xtr, ytr, 0).
local rhovec to V(2*rho, 0, 0).
local vv0x to fp * sin(rangl) / (1 - ecc * cos(rangl) )^2.
local vv0y to fp * (cos(rangl) - ecc) / (1 - ecc * cos(rangl) )^2.
local vv0 to V( vv0x, vv0y, 0 ).
local vangl to vang( vv0, V(0,1,0) ).

local dVvec to dVrem * heading(0, vangl - rangl - 0.2):vector - velocity:orbit.
local steerdir to lookdirup(dVvec, Sun:position).
lock steering to steerdir.

until vang( facing:vector, steerdir:vector ) < 0.005 and ship:angularvel:mag < 0.01 {
  set dVvec to dVrem * heading(0, vangl - rangl - 0.2):vector - velocity:orbit.
  //set steerdir to angleaxis( 360/orbit:period * dob/2, V(0,1,0) )*lookdirup(dVvec, up:vector).
  set steerdir to lookdirup(dVvec, Sun:position).
  lock steering to steerdir.
  print "Angle derivation: " + vang( facing:vector, steerdir:vector ) at (0,31).
  print "Angular speed: " + ship:angularvel:mag at (0,32).
  wait 0.01.
}
stage.

 

 

[V7 Aerospace]

Who knows? Maybe SQUAD will announce that the "KSP Challenge" is all just a big April fools joke to waste everyone's time! 

[MaverickSawyer]

On 3/29/2017 at 5:10 PM, Toastie_Buns said:

Many quickloads later:

Apologies for the small gif. Imgur decided it didn't want to upload a 50mb file.

*tockleockle*

*golf clap*

[mcirish3]

On 3/30/2017 at 2:19 PM, Hotaru said:

Snip

 

Ok fair enough!

[lajoswinkler]

OK... I need help.

The probe in Kerbin's SOI.

 

Trajectory has been set. It will impact the mohole.

 

However, even though I've set the decoupler to zero force, it releases the projectile with a push, which completely wrecks the trajectory. Not surprising, as this took cm/s adjustments.

 

My question - how do I sever the TR-2V connection without imparting any impulse on the projectile?

Edited April 2, 2017 by lajoswinkler

[0111narwhalz]

40 minutes ago, lajoswinkler said:

My question - how do I sever the TR-2V connection without imparting any impulse on the projectile?

Blowtorch staging?

[lajoswinkler]

1 hour ago, 0111narwhalz said:

Blowtorch staging?

Not only I don't have the needed docking ports, but also undocking would impart another push.

[0111narwhalz]

1 minute ago, lajoswinkler said:

Not only I don't have the needed docking ports, but also undocking would impart another push.

I didn't ask for undocking. I asked for blowtorch staging. With opposed Sepratrons.

[stibbons]

I believe the KIS/KAS mod pairing should let a Kerbal on EVA destroy parts. Maybe destroying the decoupler will separate them without any extra force?

[lajoswinkler]

29 minutes ago, 0111narwhalz said:

I didn't ask for undocking. I asked for blowtorch staging. With opposed Sepratrons.

Sepratrons? We're talking about changing the speed of the projectile in the range of milimetres per second up to few centimetres per second. A gentle nudge throws the landing spot to the equator.

 

No sepratrons.

29 minutes ago, stibbons said:

I believe the KIS/KAS mod pairing should let a Kerbal on EVA destroy parts. Maybe destroying the decoupler will separate them without any extra force?

As I recall, that gives a pretty big push in an explosion.

Maybe there's a way to edit the vessel in flight and just delete this.

[Foxster]

Umm...How about separating before setting up the shot? Hold the part in place with a frame, then split the frame and ship with decouplers when lined up, leaving the ball to go its way. 

Edited April 2, 2017 by Foxster

[Toastie_Buns]

12 hours ago, lajoswinkler said:

OK... I need help.

The probe in Kerbin's SOI.

My question - how do I sever the TR-2V connection without imparting any impulse on the projectile?

You don't. You factor in the push from the start. Technically you aren't firing the ball, at least with a push you would be. 

Edit: It's not even that hard, just put another probe up there, stage the decoupler at peri and watch the apsis rise, you'll have your m/s calculation for the actual shot. Remember, Speed = Distance / Time

Edited April 2, 2017 by Toastie_Buns

[lajoswinkler]

2 hours ago, Toastie_Buns said:

You don't. You factor in the push from the start. Technically you aren't firing the ball, at least with a push you would be. 

Edit: It's not even that hard, just put another probe up there, stage the decoupler at peri and watch the apsis rise, you'll have your m/s calculation for the actual shot. Remember, Speed = Distance / Time

I can't factor that when the target is dancing over a spherical surface so far away and my probe is at Kerbin. It took me almost an hour to set this cross to the mohole.

Also, my probe is on a hyperbolic trajectory inside Kerbin's SOI, going over 3 km/s. Each second the conditions change.

 

6 hours ago, Foxster said:

Umm...How about separating before setting up the shot? Hold the part in place with a frame, then split the frame and ship with decouplers when lined up, leaving the ball to go its way. 

Even the slightest nudge shifts the target by few kilometres. If you had a small cubical strut hitting one of the spikes at few cm/s, it would move the crash site like 5 km away.

To target the hole, I've used an ion engine throttled to 0.5 in VAB, and then the tiniest thrust in orbit. If you touch the ball from a different angle, target goes nuts. It's very frustrating. I just want the decoupler to "vanish". The probe will eventually move away from the ball by the time it reaches Moho because of the orbital drift.

Edited April 2, 2017 by lajoswinkler

[Pand5461]

@lajoswinkler, are you sure this all works without decoupling? I mean, will it still hit the hole after two SOI changes?

And, even if you delete decoupler, the orbit still may change enough to miss due to CoM shift.

[IvanSanchez]

Here's my entry. I opted for the "carefully balance some SRBs to do a perfectly vertical drop" technique plus a "why do one impactor if I can do six?" kerbalicity bonus.

 

[lajoswinkler]

1 hour ago, Pand5461 said:

@lajoswinkler, are you sure this all works without decoupling? I mean, will it still hit the hole after two SOI changes?

And, even if you delete decoupler, the orbit still may change enough to miss due to CoM shift.

I've tried it two times and it hits the same spot, some 2.4 km away from the rim. Granted, it's so difficult to tweak this by looking at the map view, but the point is that it can hit the same spot. I transition the SOIs without timewarping.

All I need to do is do few trial and errors and then quicksave when it impacts the hole. However, the challenge says the vessel needs to be uncontrollable. Maybe I should smash the antennas using a whack-a-kerbal? AFAIK it does not impart impulse, just like smashing solar panels doesn't.

[Servo]

2 hours ago, lajoswinkler said:

I can't factor that when the target is dancing over a spherical surface so far away and my probe is at Kerbin. It took me almost an hour to set this cross to the mohole.

Also, my probe is on a hyperbolic trajectory inside Kerbin's SOI, going over 3 km/s. Each second the conditions change.

 

Even the slightest nudge shifts the target by few kilometres. If you had a small cubical strut hitting one of the spikes at few cm/s, it would move the crash site like 5 km away.

To target the hole, I've used an ion engine throttled to 0.5 in VAB, and then the tiniest thrust in orbit. If you touch the ball from a different angle, target goes nuts. It's very frustrating. I just want the decoupler to "vanish". The probe will eventually move away from the ball by the time it reaches Moho because of the orbital drift.

Use opposed decouplers (one on either side of the ball), each set to zero force. Unfortunately, that will require a redesign of the craft, but it should work.

[Just Jim]

Well, they say the 13th time is the charm... 

Here's my submission.

But first, I have to say, this is not nearly as easy as it seems, and my deepest respect to those that have already done it! 

So y'all want something uncontrollable... how about we have some fun and drop a small asteroid! I won't lie... I'm not sure if it was actually my 13th attempt, but close... I had to do this several times, and tweak the maneuver just a hair each time, before I got one to drop inside. Also, I cheated a little and turned off crash damage so it wouldn't just smash into the side before I could get screenshots.

I'll put these in a spoiler thing so it doesn't eat up half a page.

Spoiler

 

 

I look forward to the next challenge!

 

[Toastie_Buns]

4 hours ago, lajoswinkler said:

I can't factor that when the target is dancing over a spherical surface so far away and my probe is at Kerbin. It took me almost an hour to set this cross to the mohole.

You don't understand me. Make a brand new probe identical to yours. Put it in low kerbin orbit. Check the apsis and peri. Fire the decoupler at peri.
Note the new apsis, work out the speed (s=d/t) by the time to apsis and distance. You will now know EXACTLY how many m/s the decoupler force is.

Now switch to your old craft, set the new maneuver to incorporate the extra m/s of the decoupler force, burn, fire the decoupler. 

VIOLA.

[lajoswinkler]

2 hours ago, Toastie_Buns said:

You don't understand me. Make a brand new probe identical to yours. Put it in low kerbin orbit. Check the apsis and peri. Fire the decoupler at peri.
Note the new apsis, work out the speed (s=d/t) by the time to apsis and distance. You will now know EXACTLY how many m/s the decoupler force is.

Now switch to your old craft, set the new maneuver to incorporate the extra m/s of the decoupler force, burn, fire the decoupler. 

VIOLA.

I'm incapable of being so precise, trust me.

4 hours ago, Servo said:

Use opposed decouplers (one on either side of the ball), each set to zero force. Unfortunately, that will require a redesign of the craft, but it should work.

Hm... But the first decoupler is attached to the probe. We aren't separating two equal masses.

probe-decoupler-projectile-decoupler

probe-decoupler     projectile     decoupler

Edited April 2, 2017 by lajoswinkler

[Foxster]

Just a thought for those putting unpowered objects into the Mohole (as was pointed out previously)...

The challenge says: "You will have to launch an uncontrolled vessel from a mothership orbiting Moho and get the vessel within Moho’s hole...". 

So, to meet the challenge's requirements, a released vessel would have to be powered to shift it from an orbital to suborbital trajectory or else de-orbited with a mass-driver arrangement firing on the mothership. 

 

Edited April 2, 2017 by Foxster

[Servo]

1 hour ago, lajoswinkler said:

I'm incapable of being so precise, trust me.

Hm... But the first decoupler is attached to the probe. We aren't separating two equal masses.

probe-decoupler-projectile-decoupler

probe-decoupler     projectile     decoupler

You are correct in that the mass decoupled won't be the same, but it's irrelevant. Some quick tests involving decouplers with different masses on them could prove this. A simple test rig could prove this:

Have a long strut in orbit, with decouplers equidistant, and facing in the same direction. Each decoupler would be set to the same staging section, and have a different mass attached. With SAS off, stage both at the same time. If the beam begins to spin, the force is inequal, if it is, then it won't spin.

Schematic:

[]-|(m=100kg)
[]
[]
[]
**
[]
[]
[]
[]-|(m=0kg)

[] = structural frame
-| = deocoupler
() = dummy mass
** = probe core

From a physics standpoint, we can analyze the system to see if there is any torque (which would indicate unbalanced force). I'll do everything in general terms .

Velocity change of the mass = Force of the decoupler * time firing / (mass of the decoupler + dummy mass) = Decoupler Impulse / (mass of the decoupler + dummy mass)

Angular momentum of the mass about the center  = m(R x V) = Length of the rod * Decoupler Impulse / (mass of the decoupler + dummy mass) * (mass of the decoupler + dummy mass) = Length of the rod * Decoupler Impulse

From this last result, we can see that the angular momentum of the mass only depends on the impulse of the decoupler and the length that it is away from the center of mass.

We can eliminate this last requirement by noting that torque is equal to L x V, or |L||V|sin(angle). If we make the force act in line with the center of mass, there will be no torque on the craft.

As angular acceleration is equal to the Torque divided by the moment of inertia (a term for mass distrubution), and there is no torque, there is no angular acceleration (spinning).

 

In the above example, there will still be linear acceleration, as both decouplers are pointed in the same direction. The change in velocity is equal to:

(Sum of net impulse) / mass.

When the two decouplers are opposed, their impulse will exactly cancel (as they are identical), leading to no linear acceleration.

 

No linear acceleration, no angular acceleration = craft is exactly on target.

 

That is, of course, assuming that Unity renders everything to enough decimal places...