Jump to content

[1.3.0] Kerbal Engineer Redux 1.1.3.0 (2017-05-28)


cybutek

Recommended Posts

  • 1 month later...
12 minutes ago, Nicky21 said:

There's a bug with reporting "suicide burn countdown".

It reports just fine in normal time, but as soon as you time warp you get very different numbers.

I trought it was just my install. Yes it happens when you warp, and I dont remember happening before.

Link to comment
Share on other sites

11 hours ago, SiCaRiO31 said:

I trought it was just my install. Yes it happens when you warp, and I dont remember happening before.

It's wrong while warping and there's another bug in normal time for low, shallow approaches where the suicide burn starts after the impact time. I reported this long ago but here's another screenshot

 

Link to comment
Share on other sites

  • 1 month later...
  • 1 month later...
  • 1 month later...
On 10/6/2021 at 7:08 AM, Krazy1 said:

It's wrong while warping and there's another bug in normal time for low, shallow approaches where the suicide burn starts after the impact time. I reported this long ago but here's another screenshot

 

I think that's a deliberate performance shortcut.

Years-old posts in the KSP subreddit claim that KER's suicide burn calculations always assume a perfectly vertical descent, which is obviously going to be wrong if you descend at an angle. This is because to get the actual downwards thrust, KER would have to poll the craft's current attitude, get the engines' thrust vectors' angle away from pointing perfectly downwards, calculate the cosine of that angle and multiply the engine thrust with that. All of which would have to recalculated every single time your craft rotated even a microscopic fraction of a degree in any direction, including from engine torque caused by your craft's weight distribution not being 100% symmetric.

I'm sure you can imagine the performance impact of doing all that every couple of milliseconds in a game that's already one of the most brutal CPU hogs I know of.

End of line, you must kill your horizontal velocity before you begin the suicide burn. That is how the suicide burn countdown will be accurate - but even so, I always seem to end up killing my descent speed before reaching the surface if I start high enough, so I think the countdown might not calculate weight loss from fuel expenditure either.

Link to comment
Share on other sites

6 hours ago, Fraktal said:

assume a perfectly vertical descent

Thanks for the info. Makes sense. It's a linear approximation and just uses the vertical component. But at least it shouldn't display an absurd value. If the suicide burn start time is after the impact time it should not display it at all. Bad data is worse than no data IMO. 

7 hours ago, Fraktal said:

doing all that every couple of milliseconds

I doubt it would be terrible. The KSP physics frame time is fixed at 20 ms. A CPU can do a COS in under a microsecond. The big performance hog is the ship part count because there are hundreds or thousands of interactions between the parts which also have springy joints. If you use the Trajectories mod, it does some fairly complex calculations like plotting the path of a ship through nonlinear atmospheric pressure. It shows the calculation time in the last tab and I'm seeing 1 ms/ 4% at the moment. I expect a suicide burn model would be similar. It could use some assumptions like a locally-flat world, constant gravity, constant mass and still be OK.  I'm not saying I have it figured out but it's seems doable.  

8 hours ago, Fraktal said:

you must kill your horizontal velocity before you begin the suicide burn. 

Then KER is right but that defeats the purpose of a suicide burn. A landing is a launch in reverse... you wouldn't launch vertical and then turn 90 deg. and try to do a horizontal burn to get to orbit, right? You could but it's not efficient.  

9 hours ago, Fraktal said:

the countdown might not calculate weight loss

I'm pretty sure it doesn't

Link to comment
Share on other sites

  • 1 month later...

On the topic of suicide burns: 

When I started maintaining KER the suicide burn did indeed only account for vertical velocity. I spent quite a bit of time in 2018 re-doing the impact/suicide calculations and adding the impact marker readout. It is a lot smarter than the previous 'vertical only' calculation, now a simulation instead of a simple equation, but it is by no means perfect. It will not account for air drag, nor weight loss during the burn. It uses approximations to estimate the angle of the craft at each step of the simulation, but as with most approximations, stray too far from the typical values and they break down. I tested it to my satisfaction using reasonably parabolic impacts and burns less than a minute or so. I did not test the calculation under on-rails warp, I would not suggest relying on the readout when that far away from the planet, nor when your horizontal velocity is on a higher order of magnitude than the vertical.

Link to comment
Share on other sites

Implementation proposal:

Additional dV and Burntimecountdown Readouts for the current stage 

  • Considdering only active engines (currently all already staged engines are considdered in the calculation, which is usefull e.g. for the far future engines from nertea, but an additional readout for only the active engines would be good for multi-engine-SSTOs, where you turn off rapier in favor of nervas.

Refinement of dV-calculation: engines, that are slightly tilted out of the prograde-direction. E.g. nerteas radial spaceX-like capsules and radial Monofuel engines.... the stock dV readouts are more precise since they do the math of calculating the part of the exhaust, that is directed prograde.

 

Edited by Rakete
Link to comment
Share on other sites

  • 3 months later...

How come I only have the ER 7500 part in the science section and not the chip as well? Both are researched and I have reinstalled the mod.

Could some other mod interfere somehow?

 

Edit: seems the mod kerbalism was the problem

Edited by mcniff
Link to comment
Share on other sites

  • 5 months later...
11 hours ago, Benjineer said:

is there a way to make your own readout/function? for example minimum pitch angle to maintain 0 vertical speed when descending/ascending from a body with low twr.

Not from the current User Interface, which exposes only those parameters available to the KSP API.  You'd have to code this computation yourself and submit a PR on GitHub.

Link to comment
Share on other sites

  • 2 months later...
  • 6 months later...

I can't detach Kerbal Engineering System (Part) and it blocks detachment of the Science Jr. that is attached to. I tried object thrower but it destroys the Science Jr as well. Is there any other soution to remove Kerbal Engineering System (Part)?

KEr.jpg

 

EDIT: I installed the KSP Part Volumes mod and now I can detach modded parts.

Edited by CanOmer
Link to comment
Share on other sites

I am sturggling a bit to understand this.  I am trying to use the land calculator, the "timer until burn start" is longer than the time until impact.  My timer also fluctuates wildly between 7 min and -300 seconds so i am really confused

Link to comment
Share on other sites

8 hours ago, fancycat said:

I am sturggling a bit to understand this.  I am trying to use the land calculator, the "timer until burn start" is longer than the time until impact.  My timer also fluctuates wildly between 7 min and -300 seconds so i am really confused

There was some discussion about this above.  It works well in most conditions but not always.

 

Link to comment
Share on other sites

On 10/29/2023 at 6:09 PM, fancycat said:

"timer until burn start" is longer than the time until impact

Time until impact assumes no intervention. The vessel will continue to accelerate by gravity and crash in that much time if you take no action.

Timer until burn start is longer because you'll be slowing down after you activate braking engines.

Edited by DeadJohn
my 2nd paragraph makes no sense; don't know what I was thinking
Link to comment
Share on other sites

I think we're talking about different things here...

10 hours ago, DeadJohn said:

Time until impact assumes no intervention. The vessel will continue to accelerate by gravity and crash in that much time if you take no action.

OK agreed.

10 hours ago, DeadJohn said:

Timer until burn start is longer because you'll be slowing down after you activate braking engines.

Then you would wait to start your suicide burn until after you crash. I think what you mean is the suicide burn length can be longer than the impact time. That is certainly possible. KER has both indicators: suicide burn length and suicide burn countdown. 

Link to comment
Share on other sites

4 hours ago, Krazy1 said:

Then you would wait to start your suicide burn until after you crash. I think what you mean is the suicide burn length can be longer than the impact time. That is certainly possible. KER has both indicators: suicide burn length and suicide burn countdown. 

Oh, yeah, you're correct. I had it backwards.

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...