schlosrat
-
Posts
600 -
Joined
-
Last visited
Content Type
Profiles
Forums
Developer Articles
KSP2 Release Notes
Bug Reports
Status Replies posted by schlosrat
-
Hello,
as you suggested, I took a brief look at FlightPlan.
I guess everything below this line:
I took a brief look a FlightPlan.
I suppose everything below this line: https://github.com/schlosrat/FlightPlan/blob/c0431bb3ce4c18de62e2ac74367d26db09a395a5/FlightPlanProject/FlightPlanPlugin.cs#L956
Is up for external usage.It should be pretty straight forward to create bindings for those functions. The hardest part is probably to make that an optional dependency.
-
If you look at the README.md you'll find info on how to setup FP as a soft dependency so you can check to see if it's there and then if it is you can call it's public methods via reflection methods.
Also, if FP is loaded, then Node Manager should also be loaded and it's methods can be called similarly.
Also, both are available as nuget packages, so you can access them that way just like SpaceWarp and BepInEx so you don't even need to have their DLLs in your external_dlls folder.
-
-
Hello,
as you suggested, I took a brief look at FlightPlan.
I guess everything below this line:
I took a brief look a FlightPlan.
I suppose everything below this line: https://github.com/schlosrat/FlightPlan/blob/c0431bb3ce4c18de62e2ac74367d26db09a395a5/FlightPlanProject/FlightPlanPlugin.cs#L956
Is up for external usage.It should be pretty straight forward to create bindings for those functions. The hardest part is probably to make that an optional dependency.
-
Please could you read the private message I sent you via the forum messenger system?
I'd like to discuss vessels and turn shapes and stuff, if you're interested
But not in the MechJeb thread, it's a bit off topic.
For now I do have some insights on your vessel and your findings, if you're interested?
Your vessel's TWR at launch isn't 1.51, it's 1.36. That's because of your engines having lower efficiency and lower thrust at low altitudes. (In the VSB using mechjeb you can see that as SLT.)
The messier bit is understanding what your TWR is after your first stage. I tend to edit the Vessel Info and turn on "Local TWR". This shows the actual TWR at the current altitude. (The Delta-V status show the TWR [assumes you're in space] or the SLT [assumes you're at sea level].)
When I run your vessel with 80% turn I can see the 1.36 "Local TWR" at launch. When it stages the boosters away that drops to 1.24 "Local TWR".
That 1.24 TWR sustainer is why you need 80% turn shape, it's actually a very low TWR at that early point in the ascent ("should" be up to ~1.6 by then). This means that during your gravity turn the vessel is "falling over due to gravity" quite quickly. Even with 80% turn you can see that, while MechJeb is pointing at target, Pro-Grade is constantly lower than target; MechJeb is effectively constantly trying to point the vessel back up, trying to counteract the vessel's natural gravity turn; in order to fly the planned shape. With a shallower turn (such as the 40% everyone wants you to use) it would waste even more fuel trying to counteract the natural rate of turn of the vessel (due to its low TWR).
If I take out the asparagus your vessel keeps a higher TWR for a lot longer. In this configuration the best turn shape is about 55% (I didn't check very rigorously; 1191dv @ 40%, 1210dv @ 45%, 1224dv @ 50%, 1224dv @ 55%, 1219dv @ 60%)
What does this show me? That by using asparagus you've made a very capable low TWR launch vehicle that indeed benefits from a steeper turn. (The benefits of the asparagus are higher than the benefits of a shallower ascent.)
I'd love to carry this time of chatter on, if you would too, please let me know?
-
Please could you read the private message I sent you via the forum messenger system?
I'd like to discuss vessels and turn shapes and stuff, if you're interested
But not in the MechJeb thread, it's a bit off topic.
For now I do have some insights on your vessel and your findings, if you're interested?
Your vessel's TWR at launch isn't 1.51, it's 1.36. That's because of your engines having lower efficiency and lower thrust at low altitudes. (In the VSB using mechjeb you can see that as SLT.)
The messier bit is understanding what your TWR is after your first stage. I tend to edit the Vessel Info and turn on "Local TWR". This shows the actual TWR at the current altitude. (The Delta-V status show the TWR [assumes you're in space] or the SLT [assumes you're at sea level].)
When I run your vessel with 80% turn I can see the 1.36 "Local TWR" at launch. When it stages the boosters away that drops to 1.24 "Local TWR".
That 1.24 TWR sustainer is why you need 80% turn shape, it's actually a very low TWR at that early point in the ascent ("should" be up to ~1.6 by then). This means that during your gravity turn the vessel is "falling over due to gravity" quite quickly. Even with 80% turn you can see that, while MechJeb is pointing at target, Pro-Grade is constantly lower than target; MechJeb is effectively constantly trying to point the vessel back up, trying to counteract the vessel's natural gravity turn; in order to fly the planned shape. With a shallower turn (such as the 40% everyone wants you to use) it would waste even more fuel trying to counteract the natural rate of turn of the vessel (due to its low TWR).
If I take out the asparagus your vessel keeps a higher TWR for a lot longer. In this configuration the best turn shape is about 55% (I didn't check very rigorously; 1191dv @ 40%, 1210dv @ 45%, 1224dv @ 50%, 1224dv @ 55%, 1219dv @ 60%)
What does this show me? That by using asparagus you've made a very capable low TWR launch vehicle that indeed benefits from a steeper turn. (The benefits of the asparagus are higher than the benefits of a shallower ascent.)
I'd love to carry this time of chatter on, if you would too, please let me know?
-
Well now I can say I've done my first MechJeb assisted manual launch to orbit!
Not wanting to supply a pilot (all of mine are busy on other missions) I opted to slap in a probe core between the capsule and the claw so that I could make use of the stock SAS. I'm not sure if I really needed that now though and may try this again...
Once I figured out the rather obvious step that I didn't want to turn on the MJ Autopilot, but I DID need to have the other button above that selected to show nav ball guidance, the rest was fairly simple.
I think I did a bit of a sloppy initial turn, but after that everything went smoothly and I arrived in a 100.720 x 100.591 orbit at 0.489 inclination with 1811 m/s left in the tank. The fairing was the only part to really get hot, and it never even came close to it's limits. I didn't bother to resize the fairing, so the claw was protruding slightly from it.
I'm going to try it again with the MechJeb SmartSAS in place of the SAS provided by the probe core and see if I can get that to work.
BTW, I tried inserting the CFG file that's supposed to put MJ functionality in every capsule, but I must have done that wrong as it doesn't work for me. All I did was copy the code from the version that's supposed to be compatible with science and career modes and place it in a CFG file in the GameData folder. I already have MM so I'm not sure what I did wrong there.
-
-
Please could you read the private message I sent you via the forum messenger system?
I'd like to discuss vessels and turn shapes and stuff, if you're interested
But not in the MechJeb thread, it's a bit off topic.
For now I do have some insights on your vessel and your findings, if you're interested?
Your vessel's TWR at launch isn't 1.51, it's 1.36. That's because of your engines having lower efficiency and lower thrust at low altitudes. (In the VSB using mechjeb you can see that as SLT.)
The messier bit is understanding what your TWR is after your first stage. I tend to edit the Vessel Info and turn on "Local TWR". This shows the actual TWR at the current altitude. (The Delta-V status show the TWR [assumes you're in space] or the SLT [assumes you're at sea level].)
When I run your vessel with 80% turn I can see the 1.36 "Local TWR" at launch. When it stages the boosters away that drops to 1.24 "Local TWR".
That 1.24 TWR sustainer is why you need 80% turn shape, it's actually a very low TWR at that early point in the ascent ("should" be up to ~1.6 by then). This means that during your gravity turn the vessel is "falling over due to gravity" quite quickly. Even with 80% turn you can see that, while MechJeb is pointing at target, Pro-Grade is constantly lower than target; MechJeb is effectively constantly trying to point the vessel back up, trying to counteract the vessel's natural gravity turn; in order to fly the planned shape. With a shallower turn (such as the 40% everyone wants you to use) it would waste even more fuel trying to counteract the natural rate of turn of the vessel (due to its low TWR).
If I take out the asparagus your vessel keeps a higher TWR for a lot longer. In this configuration the best turn shape is about 55% (I didn't check very rigorously; 1191dv @ 40%, 1210dv @ 45%, 1224dv @ 50%, 1224dv @ 55%, 1219dv @ 60%)
What does this show me? That by using asparagus you've made a very capable low TWR launch vehicle that indeed benefits from a steeper turn. (The benefits of the asparagus are higher than the benefits of a shallower ascent.)
I'd love to carry this time of chatter on, if you would too, please let me know?
-
One of the things that really strikes me about the results you've found is that the formula for a better ascent is actually quite simple. You summed it up nicely in this last post. I can't imagine why MJ would not be able to be programmed to do just what you've described in the semi-manual mode.
Regarding how to pick a better turn shape, I lean toward the notion you suggested as well: assume good aerodynamics. But why not follow the Launch to Rendezvous paradigm? Let the game be the simulation that produces the numbers that MechJeb needs. All it needs is to get a reasonably decent start and iterate from there. If the assumption of good aerodynamics, or simply no drag losses, can get you close enough that you can make it to orbit (albeit inefficiently), then the flight up will give you the data needed to refine to a better turn shape. The next flight starts with the refined shape, and using the resulting performance incrementally moves closer to the optimum. The more you use any one design, the better the estimate of the optimal turn shape. The only thing needed is a way to take the data from the flight and use that to point in the direction the turn shape need to change for better performance. Kind of like the Newton method for curve fitting: https://en.wikipedia.org/wiki/Gauss–Newton_algorithm. Essentially the problem that is being solved is to minimize losses. I'm assuming drag losses and gravity losses are equally costly.
But given what you've found about semi-manual flights I'm also wondering how that would impact it. It could make it even simpler since you're only guiding to the target for the first part of the flight and thereafter you're guiding to prograde. So the turn shape really only applies until the resulting target aligns with the prograde vector, and thereafter the optimal solution is assumed to be fly prograde as hard as you can - but don't blow up.
Your thoughts?
-
-
Please could you read the private message I sent you via the forum messenger system?
I'd like to discuss vessels and turn shapes and stuff, if you're interested
But not in the MechJeb thread, it's a bit off topic.
For now I do have some insights on your vessel and your findings, if you're interested?
Your vessel's TWR at launch isn't 1.51, it's 1.36. That's because of your engines having lower efficiency and lower thrust at low altitudes. (In the VSB using mechjeb you can see that as SLT.)
The messier bit is understanding what your TWR is after your first stage. I tend to edit the Vessel Info and turn on "Local TWR". This shows the actual TWR at the current altitude. (The Delta-V status show the TWR [assumes you're in space] or the SLT [assumes you're at sea level].)
When I run your vessel with 80% turn I can see the 1.36 "Local TWR" at launch. When it stages the boosters away that drops to 1.24 "Local TWR".
That 1.24 TWR sustainer is why you need 80% turn shape, it's actually a very low TWR at that early point in the ascent ("should" be up to ~1.6 by then). This means that during your gravity turn the vessel is "falling over due to gravity" quite quickly. Even with 80% turn you can see that, while MechJeb is pointing at target, Pro-Grade is constantly lower than target; MechJeb is effectively constantly trying to point the vessel back up, trying to counteract the vessel's natural gravity turn; in order to fly the planned shape. With a shallower turn (such as the 40% everyone wants you to use) it would waste even more fuel trying to counteract the natural rate of turn of the vessel (due to its low TWR).
If I take out the asparagus your vessel keeps a higher TWR for a lot longer. In this configuration the best turn shape is about 55% (I didn't check very rigorously; 1191dv @ 40%, 1210dv @ 45%, 1224dv @ 50%, 1224dv @ 55%, 1219dv @ 60%)
What does this show me? That by using asparagus you've made a very capable low TWR launch vehicle that indeed benefits from a steeper turn. (The benefits of the asparagus are higher than the benefits of a shallower ascent.)
I'd love to carry this time of chatter on, if you would too, please let me know?
-
Outstanding! I must say I'm intrigued by what you told me so far. I'm going to have to look into kOS and play around with that a bit.
One thought I've got goes back to the question of can MJ pick an estimated turn shape which would be reasonable for a given rocket. Clearly it can estimate the TWR of the craft over time, bug maybe it wouldn't have a way of knowing what altitude you'd be at over time without actually flying the craft. If it could come up with even a crude estimate for altitude over time, the it could probably get to an estimate of speed over time and from there get to an estimate of drag over time. I'm imagining a scenario in which, much like launch to rendezvous, it starts you out with a crude but safe estimate, then you launch to orbit, the you revert back to launch, and from there it knows what the turn shape should actually be.
-
-
Please could you read the private message I sent you via the forum messenger system?
I'd like to discuss vessels and turn shapes and stuff, if you're interested
But not in the MechJeb thread, it's a bit off topic.
For now I do have some insights on your vessel and your findings, if you're interested?
Your vessel's TWR at launch isn't 1.51, it's 1.36. That's because of your engines having lower efficiency and lower thrust at low altitudes. (In the VSB using mechjeb you can see that as SLT.)
The messier bit is understanding what your TWR is after your first stage. I tend to edit the Vessel Info and turn on "Local TWR". This shows the actual TWR at the current altitude. (The Delta-V status show the TWR [assumes you're in space] or the SLT [assumes you're at sea level].)
When I run your vessel with 80% turn I can see the 1.36 "Local TWR" at launch. When it stages the boosters away that drops to 1.24 "Local TWR".
That 1.24 TWR sustainer is why you need 80% turn shape, it's actually a very low TWR at that early point in the ascent ("should" be up to ~1.6 by then). This means that during your gravity turn the vessel is "falling over due to gravity" quite quickly. Even with 80% turn you can see that, while MechJeb is pointing at target, Pro-Grade is constantly lower than target; MechJeb is effectively constantly trying to point the vessel back up, trying to counteract the vessel's natural gravity turn; in order to fly the planned shape. With a shallower turn (such as the 40% everyone wants you to use) it would waste even more fuel trying to counteract the natural rate of turn of the vessel (due to its low TWR).
If I take out the asparagus your vessel keeps a higher TWR for a lot longer. In this configuration the best turn shape is about 55% (I didn't check very rigorously; 1191dv @ 40%, 1210dv @ 45%, 1224dv @ 50%, 1224dv @ 55%, 1219dv @ 60%)
What does this show me? That by using asparagus you've made a very capable low TWR launch vehicle that indeed benefits from a steeper turn. (The benefits of the asparagus are higher than the benefits of a shallower ascent.)
I'd love to carry this time of chatter on, if you would too, please let me know?
