Jump to content

[PART, 1.0.2] Anatid Robotics / MuMech - MechJeb - Autopilot - Historical thread


r4m0n

Recommended Posts

I've got a quick question. I have a target selected in ascent guidance and when I click either "launch to rendezvous" or "launch into plane of target" I get a countdown and time warp, but then nothing. It will count down to zero seconds, then just sit there doing nothing. Both buttons have this affect. What am I doing wrong here? Thank you.

Did you click Enable Autopilot? LTR will time warp even without the autopilot being activated.

Link to comment
Share on other sites

As if you guys don't have enough to do ... ( :) ), here is a suggestion for the "Transfer to Another Planet" option in the Maneuver Planner (and if this function is already available I didn't know it, I apologize).

For large vessels, especially those using nuclear engines (e.g., low thrust) the transfer burn can easily be >10 minutes. I had one last night that was pushing 12 and I've heard of others with 20+ minute transfer burns. If transferring from Kerbin (especially lower orbits), the length of the burn can easily take up a sizable arc-length of an orbit, making it very inefficient as the craft "chases" the target marker around the navball due to the orbit.

Would it be possible (is it already?) to have MJ take a look at the estimated burn length and enable an option to break up the burn into shorter segments and perform them in 5 - 6 minute long chunks at periapsis during successive orbits? This would allow greater efficiency from the Oberth effect and frankly, most planetary transfer windows aren't so tight that an hour or three to make "segmented" transfer burns like this will affect things very much. Besides which, most people plot a fine-tuning burn once out of the SOI of the starting point for the transfer anyway.

Link to comment
Share on other sites

I had the some thought while watching the latest Scott Manley video. And that would help for ion engine powered ship too.

I'll think about it, but unless I find a quick/easy way to do it don't expect to see that soon.

Link to comment
Share on other sites

Deadly Reentry does not mess with anything MJ related, it works fine. And he is working on FAR support, check the link on sarbian's sig about Optional MJ Modules

Link to comment
Share on other sites

It depends on what you mean bat that.

Some FAR support is done with the additional module linked in my signature. I am kinda stuck on how to implement the other required part (IE the landing AP) since the way FAR works I can't do a threaded simulation (and don't go asking ferram for a fix, his code is great). I have ideas but I need time to test and implement them and right now I'd rather fix bug that impact everyone in MJ.

DRE support ? What is needed for DRE ?

Link to comment
Share on other sites

Is there a target distance threshold under which the rendezvous autopilot will attempt small burns to close the separation, or will it always perform the same procedure of creating a new phasing orbit and then intercepting? Sometimes I will launch to rendezvous and arrive within 10 km of the target, but even then it has to do the whole routine and it will take about 2 hours game time to reach the next intercept.

Link to comment
Share on other sites

I had the some thought while watching the latest Scott Manley video. And that would help for ion engine powered ship too.

I'll think about it, but unless I find a quick/easy way to do it don't expect to see that soon.

If I were to add anything to Mechjeb it would be things like this. I have been flying with ions a lot recently, and some sort of burn automation I'd essential for a largeish ion craft. I also refuse to use nukes, so often my large intetplantary craft have a single, high ISP engine to save launch mass, so I often have TWR<0.1, which can mean splitting burns over multiple orbits. It is no big deal, especially if you start a few days before you need to be on your way. I have a series of spreadsheets for calcing things like raising AP over multiple orbits, or planning to be at the PE at a certain time by raising th AP to adjust the orbital period. That could be useful to come back to do the rest of the transfer burn at exactly the right time, or planning the time to AP to get into the correct geosync orbit.

If I do start to add any such features to my fork of Mechjeb, I will consult with others and offer it to be pulled once it works.

Link to comment
Share on other sites

Bobe : Looking at the code it should be if the distance to target is lower than the controlled ship orbit Radius / 25

And by radius you mean altitude right? Honestly, I do get confused sometimes because the periapsis and apoapsis are always in terms of altitude, rather than the distance to the orbital focus.

Looks like that would return the distance in metres as opposed to kilometres. So an orbit with an altitude of 100,000 km would return a threshold of 4 km, which seems about right.

Link to comment
Share on other sites

No it's the real radius. The code define it as the distance from the ship center of mass to the center of the body you orbit around.

There is a test with the semi major axis too but it's not relevant to your need. You can take a look at the code, it's not the most complex part of MJ and it's well commented : https://github.com/MuMech/MechJeb2/blob/master/MechJeb2/MechJebModuleRendezvousAutopilot.cs#L68

Mostly ensuring the sanity (as defined by DRE) of your re-entry path I'd imagine.

I could display some useful info for DRE but I doubt I'll implement a way to compute the "best" re entry path. But if someone submit a patch I ll welcome it.

Edited by sarbian
Link to comment
Share on other sites

No it's the real radius. The code define it as the distance from the ship center of mass to the center of the body you orbit around.

There is a test with the semi major axis too but it's not relevant to your need. You can take a look at the code, it's not the most complex part of MJ and it's well commented : https://github.com/MuMech/MechJeb2/blob/master/MechJeb2/MechJebModuleRendezvousAutopilot.cs#L68

In that case it still seems to make sense, because I just launched to a distance of 20 km (28 km threshold) and is now performing the correction burns instead of the phasing burns. However, I have noticed occasions where it still did the phasing burns inside that threshold, but I'll just chalk that up to MJ's slight bugginess.

Link to comment
Share on other sites

I had the some thought while watching the latest Scott Manley video. And that would help for ion engine powered ship too.

I'll think about it, but unless I find a quick/easy way to do it don't expect to see that soon.

Think about how you'd deal with the same situation if you were piloting. You'd set up multiple waypoints with the total DV spread across the. Each waypoint should be at the same orbital position as the singular waypoint created the way MJ is now. Put another way, all waypoints after the first would be at periapsis.

When I'm dealing with MJ in that situation I set up that one waypoint and manually do successive burns (using +node in smartass), each terminated when I judge the burn to be too inefficient as I pass the waypoint's position.

It also occurs to me that I might also use the waypoint editor to shorten the initial node then create additional multiple waypoint nodes (as before, each after the first snapped to periapsis)

Link to comment
Share on other sites

I could display some useful info for DRE but I doubt I'll implement a way to compute the "best" re entry path. But if someone submit a patch I ll welcome it.

I suspect you're doing the bulk of the calculations required since you're already handling computing max drag. How is that being determined? And didn't there also used to be a calculation for max g-forces or did I imagine that?

Link to comment
Share on other sites

I suspect you're doing the bulk of the calculations required since you're already handling computing max drag. How is that being determined? And didn't there also used to be a calculation for max g-forces or did I imagine that?

Yes, both are computed by the simulation.

max g-forces should be displayed as "Max drag" when you are on re entry trajectory.

Link to comment
Share on other sites

Is it normal that the excepted landing site marker isn't visible if the landing site is on water?

(it's the case for Kerbin but I didn't tested it for other celestial body with water).

the marker is placed on the terrain. it cannot be seen through the water so, if the terrain is under water then no, it wont be visible.

Link to comment
Share on other sites

I don't why they would be incompatible. There may be a window id conflict. Did you see errors in the log ?

And beside, why do you need void + MJ ? They more or less display the same think AFAIK (beside the nice VOID hud)

Link to comment
Share on other sites

I'll check the logs, the reason I was looking at both was for KSP interstellar relay network setup. I'm having trouble with my relays drifting after a hundred days or so, despite my best efforts at getting them aligned. VOID displays just a bit more information in that regard, though I think at this point I'm just going to get my relays into reasonable positions and then hyperedit them into the right spot.

Link to comment
Share on other sites

Open the attitude adjustment windows and uncheck the Auto Tune Tf. It seems to be a bad idea in atmosphere ...

I'm having a similar problem to BARCLONE, I think. Rockets that used to fly perfectly in the previous version now wander all over the sky at some point during boost. Also, when trying to point the ship while in orbit (to align with a node or do a landing at target burn) MechJeb's pointing wanders around crazily like a blind man trying to find the node with his white cane. I can easily point it under manual control, so there is plenty of control authority on the ship.

I looked in the Attitude Adjustment window, but I could find no "Auto Tune Tf" option to uncheck. I unchecked the "Use stock SAS" box (the only option in the window), and this appears to have solved the problem (subject to further tests).

(I'm using MecjJeb 2.1.1 on a Macintosh running OS X 10.8.5)

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...