Youen

[old thread] Trajectories : atmospheric predictions

Recommended Posts

I've a request for this awesome mod. I don't know what the name would be, but could you add a "body fixed" mode where instead of drawing the orbit, you draw on the planet, where you'd be over it during your orbit?

This would aid in plotting orbits with the pinpoint scanner to make sure it's passing directly over a deposit.

AKA I suck at figuring out perspective.

Share this post


Link to post
Share on other sites
I've a request for this awesome mod. I don't know what the name would be, but could you add a "body fixed" mode where instead of drawing the orbit, you draw on the planet, where you'd be over it during your orbit?

This would aid in plotting orbits with the pinpoint scanner to make sure it's passing directly over a deposit.

AKA I suck at figuring out perspective.

I suppose you've already seen the existing "body fixed mode" in the Trajectories UI window? When enabled, it plots the trajectory relatively to the body, but it is displayed in space (in 3D).

If I understand you correctly, what you want is the same thing, but projected on the planet surface instead? To check if the trajectory gets exactly above a point on the ground, you could align the camera with the center of the body (it is displayed if you zoom out a bit), and the point you want. Then you'll see if the 3D trajectory is above it or not. I'm not sure if projecting the trajectory on the ground would really help or just clutter the interface a bit more.

- - - Updated - - -

The debug messages are still on! Otherwise seems to work fine.

Oups :blush: I'll fix that asap.

Share this post


Link to post
Share on other sites

I was trying to aerobrake at Eve and Duna and the predicted trajectory was wrong. When I was flying through the atmosphere, the white line was changing and final AP was lower than it was predicted. Yes, I set the AoA option to retrograde and was following it through my flight. FAR is installed.

Share this post


Link to post
Share on other sites

Besides the debug messages I'd like to ask if there is a way to read out the apoapsis of the predicted trajectory?

Share this post


Link to post
Share on other sites

Hello there.

I've had some trouble with this mod. Each time I have a re-entry trajectory (around Kerbin), the mod seems the nicely calculate the atmospheric path. However, as soon as I hit the atmosphere, the red trajectory and impact point starts to move, shortening the path over the planet, making the spacecraft impact point moving closer to the point of atmospheric entry. There are no wings or aero-brake modules. I also re-enter retrograde, not forgetting to tick the option "retrograde" in the mod's UI. When I re-enter from a highly elliptic orbit, this effect is much amplified. In some cases shifting the expected impact point a quarter of the planet from it's initial point. Also, the effect that user Peaceduke experienced with aero-braking manoeuvres around Eve and Duna seems to be of the same nature I am experiencing here. I could even replicate his experience around kerbin with shifting calculated orbits.

What is causing these anomalies? Is it a problem with the new stock aerodynamics model that KSP uses? Do I have a faulty setting somewhere? Is there a fundamental problem with the mod?

Oh, I did use quite a few mods. Yet when I uninstalled them all except the Trajectories mod, and I still have the same problems.

Share this post


Link to post
Share on other sites
I was trying to aerobrake at Eve and Duna and the predicted trajectory was wrong. When I was flying through the atmosphere, the white line was changing and final AP was lower than it was predicted. Yes, I set the AoA option to retrograde and was following it through my flight. FAR is installed.

I dont think this has the maths in this for FAR just yet, only stock?

Share this post


Link to post
Share on other sites
I dont think this has the maths in this for FAR just yet, only stock?

The setting window says: "Aerodynamic model: FAR". Maybe it detects that I have FAR installed, but still uses stock model.

Share this post


Link to post
Share on other sites
The setting window says: "Aerodynamic model: FAR". Maybe it detects that I have FAR installed, but still uses stock model.

How can it be though, that even though I use stock aerodynamics, that I have the same issues as you? (also see my previous post)

Share this post


Link to post
Share on other sites
I've a request for this awesome mod. I don't know what the name would be, but could you add a "body fixed" mode where instead of drawing the orbit, you draw on the planet, where you'd be over it during your orbit?

This would aid in plotting orbits with the pinpoint scanner to make sure it's passing directly over a deposit.

AKA I suck at figuring out perspective.

I think what you are requesting is called a ground track.

Share this post


Link to post
Share on other sites
Hello there.

I've had some trouble with this mod. Each time I have a re-entry trajectory (around Kerbin), the mod seems the nicely calculate the atmospheric path. However, as soon as I hit the atmosphere, the red trajectory and impact point starts to move, shortening the path over the planet, making the spacecraft impact point moving closer to the point of atmospheric entry. There are no wings or aero-brake modules. I also re-enter retrograde, not forgetting to tick the option "retrograde" in the mod's UI. When I re-enter from a highly elliptic orbit, this effect is much amplified. In some cases shifting the expected impact point a quarter of the planet from it's initial point. Also, the effect that user Peaceduke experienced with aero-braking manoeuvres around Eve and Duna seems to be of the same nature I am experiencing here. I could even replicate his experience around kerbin with shifting calculated orbits.

What is causing these anomalies? Is it a problem with the new stock aerodynamics model that KSP uses? Do I have a faulty setting somewhere? Is there a fundamental problem with the mod?

Oh, I did use quite a few mods. Yet when I uninstalled them all except the Trajectories mod, and I still have the same problems.

Just to make sure, did you use the latest version (1.4)? If not, try clicking the "update now" button just before reentry (this step should not be needed with 1.4)

Also, can you post your .craft file so that I can check if I can reproduce the issue (with only stock parts would make my life easier)? Thanks.

Edited by Youen

Share this post


Link to post
Share on other sites
Just to make sure, did you use the latest version (1.4)? If not, try clicking the "update now" button just before reentry (this step should not be needed with 1.4)

Also, can you post your .craft file so that I can check if I can reproduce the issue (with only stock parts would make my life easier)? Thanks.

Thank you for taking note of my post.

I am in fact using the latest version of KSP, and I've trying the "update now" feature of the mod many a times. Here is the craft file.

For testing I've left the last rocket engine stage on the ship at re-entry. Even though this makes the ship unstable and tend to want to flip around, the effect of chancing landing spot dramatically happens well before this flip out occurs.

Maybe the problem is caused by the fact that I'm returning from such a highly eliptical orbit? the small variations in speed making a big difference due to the high orbit? I'm at a loss.

Share this post


Link to post
Share on other sites

I haven't investigated further yet, but I can confirm there is a huge difference between the predicted trajectory and the actual one with your ship. You're doing nothing wrong, it's a bug. I'll keep you posted if I find the cause (it's probably related to one or more of the parts you have on this vessel that are not correctly predicted).

- - - Updated - - -

OK, it's the FL-A10 and FL-A5 adapters (especially the FL-A5 is predicted to have drag much lower than what actually happens). I don't know yet why these parts don't work as expected.

- - - Updated - - -

I've fixed the bug and will published a release soon. These parts have the flag "PhysicsSignificance" set to 1 in the config file. I don't know exactly what that means, but it seems the effect is that they don't have an associated rigid body in the game, and the mod was ignoring drag on parts which don't have a rigid body. So the drag produced by these parts was simply ignored by the prediction, which cause the observed behavior.

EDIT: it's actually more subtle than that. Parts have two members referencing a rigid body: Part.rb and Part.Rigidbody. One could think they are the same, but apparently sometimes Part.rb is null while Part.Rigidbody isn't. The correct one to test in this case seems to be Part.Rigidbody.

Edited by Youen

Share this post


Link to post
Share on other sites
I've fixed the bug and will published a release soon. These parts have the flag "PhysicsSignificance" set to 1 in the config file. I don't know exactly what that means, but it seems the effect is that they don't have an associated rigid body in the game, and the mod was ignoring drag on parts which don't have a rigid body. So the drag produced by these parts was simply ignored by the prediction, which cause the observed behavior.

EDIT: it's actually more subtle than that. Parts have two members referencing a rigid body: Part.rb and Part.Rigidbody. One could think they are the same, but apparently sometimes Part.rb is null while Part.Rigidbody isn't. The correct one to test in this case seems to be Part.Rigidbody.

Splendid! I'm really impressed with your dedication to the mod. And I'm happy that I got to help refine and polish the mod with simply playing the game using the mod. I'll continue making new and unusual designs so I might filter out some more buggy features.

Looking forward to the updated mod. Keep up the good work! amazing! :D

Share this post


Link to post
Share on other sites

I've also fixed the following bugs: air density was not completely accurate (due to temperature changing for different latitudes), and mass of "physicsless" parts was not taken in account (they do not have drag, but still have a mass). I'm now quite confident that results will be accurate for the stock model (keep in mind that some situations are intrinsincly inaccurate, such as very long aerobrake trajectories, unstable crafts that can't keep the desired orientation, etc.).

Version 1.4.2 is released on github and kerbal stuff.

Share this post


Link to post
Share on other sites

Updated to 1.4.2 and now it doesn't display anything. In the window, when I hit "Display trajectory", the "error(s)" counter in the Settings dropdown just climbs like an altimeter. No trajectories are displayed, nothing. As if nothing were installed.

EDIT: It appears to be just one craft of my in orbit of Kerbin; it was on a suborbital aerobrake trajectory from the save. Tried lifting its peri out of atmo and lowering it again. Still nothing. Error counter under settings is now in the 20,000 range.

Edited by Shattermage

Share this post


Link to post
Share on other sites
Updated to 1.4.2 and now it doesn't display anything. In the window, when I hit "Display trajectory", the "error(s)" counter in the Settings dropdown just climbs like an altimeter. No trajectories are displayed, nothing. As if nothing were installed.

EDIT: It appears to be just one craft of my in orbit of Kerbin; it was on a suborbital aerobrake trajectory from the save. Tried lifting its peri out of atmo and lowering it again. Still nothing. Error counter under settings is now in the 20,000 range.

Does your ship contain struts or fuel ducts? I just made a test where I get errors with those (and only if FAR is not installed).

Seems to be because they don't use a drag cube like other parts, but the spherical model. I'll fix that in the next release. Also, they were probably completely ignored in previous versions, which explains why you only have the bug now.

Edited by Youen

Share this post


Link to post
Share on other sites

Hi,

On the wiki/Physicsless_part we can read this:

The mass of the part and the drag experienced by it are added to the parent's part.

I'm nearly sure you already know this, but maybe there is a chance it will help you for this great mod.

What i understand from your last posts: physicless part's mass fixed, but not yet the drag.

Edited by Skalou

Share this post


Link to post
Share on other sites
Does your ship contain struts or fuel ducts? I just made a test where I get errors with those (and only if FAR is not installed).

Seems to be because they don't use a drag cube like other parts, but the spherical model. I'll fix that in the next release. Also, they were probably completely ignored in previous versions, which explains why you only have the bug now.

Oh, you know what... Yes. I have 4 girder segments and a few struts on it; that's gotta be it. No worries now as I managed to safely land it about 70km north of KSC last night in the Highlands. Thanks.

Share this post


Link to post
Share on other sites

@Skalou: thanks for the link, I didn't read that yet. But it seems to work correctly since 1.4.2, a vessel made of a probe core covered with tens of RCS blocks is correctly predicted

@Shattermage: the bug is fixed in version 1.4.3 that I just released

Share this post


Link to post
Share on other sites

I see no reason why it wouldn't work, but I haven't tested. Let me know if you have issues.

Share this post


Link to post
Share on other sites

Is it correct that this mod only works on planets with an atmosphere? I've only seen it at work on Kerbin, and it doesn't seem to work on the Mun or Minmus.

Share this post


Link to post
Share on other sites
Is it correct that this mod only works on planets with an atmosphere? I've only seen it at work on Kerbin, and it doesn't seem to work on the Mun or Minmus.

It works better - in my experience - with airless planets and moons. When you're ready to land, switch to "Body Fixed Mode" and it'll always show you exactly where you'll land and the path you'll take over the surface. Makes nailing bases and waypoints from odd orbits a breeze.

Share this post


Link to post
Share on other sites
Is it correct that this mod only works on planets with an atmosphere? I've only seen it at work on Kerbin, and it doesn't seem to work on the Mun or Minmus.

It works on all bodies, but the landing X is sometimes obscured under the surface (especially on cratered and mountainous bodies like the Mun). Like 5thHorseman suggests, using body-fixed mode will show you where you will land; alternatively/also, if you set the target to "current impact", the green X is usually more visible.

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.