Munar Excursion Module Needs Tweaked

[Caelib]

I appreciate this being added to the game, but the RCS + SAS is so unstable when docking it's virtually worthless (this is even with RCS toned down via CAPSLOCK).  Basically, this module is NEVER going be balanced with the placement of RCS thrusters after ascent from the surface of Mun to an orbiter ... so it just spins out of control because there's no reaction wheel.  In addition, simply performing an EVA in orbit sends it into a spin even when there's a pilot inside holding with SAS and RCS enabled.

It took me a grand total of one attempt to figure out all of this so I am not sure who at Squad tested this and passed it.

Edited March 15, 2018 by Caelib

[ZooNamedGames]

Well, to be realistic- the LM never docked with the CSM during Apollo- the CSM did. So this over control is actually reasonable seeing as you're never intended to perform docking operations with the LM, but rather with the CSM. The LM only rendezvoused with the CSM, once the relative velocity was zeroed at a decent distance, the CSM would take over the closing approach and the final actual docking with the LM, with the LM only adjusting to stay aligned with the CSM. So maybe a change in operation may fix your problems?

That said, no reaction wheel is also realistic. Real manned spacecraft usually do not have the spare weight or fuel to warrant bringing them along. You can also pack one on the part but in my experience, this is a decent remake of the LM (minus the fact IVA is worthless since you're sitting down and away from the window, I mean come on gentlemen, the small windows were because the astronauts stood at the window to see).

Best suggestion is to try shifting things up and see how they fair slightly differently. If problems continue, then we'll address them individually.

Edited March 16, 2018 by ZooNamedGames

[sevenperforce]

Also on the subject of realism...EVA in orbit was...not really a thing for the LM.

[fourfa]

Try this, for shifting the CoM to actually be inside the craft (LOL), which will fix a lot of what the OP is concerned about:

Now - the rotation on EVA is a variation on the traditional KSP 'ladder drive.'  On egress, the newly spawned Kerbal helmet collider interferes with the horizontal shelf above the MEM hatch.  The collision solver interprets that as a force pushing the two apart, and presto, rotation.  Might also be fixed by adjusting the model so the EVA spawn point is lower or farther out.

[ZooNamedGames]

3 hours ago, fourfa said:

Try this, for shifting the CoM to actually be inside the craft (LOL), which will fix a lot of what the OP is concerned about:

Now - the rotation on EVA is a variation on the traditional KSP 'ladder drive.'  On egress, the newly spawned Kerbal helmet collider interferes with the horizontal shelf above the MEM hatch.  The collision solver interprets that as a force pushing the two apart, and presto, rotation.  Might also be fixed by adjusting the model so the EVA spawn point is lower or farther out.

I figured that might be the case but I wanted someone more aware of the issue to discuss it first.

[DerekL1963]

5 hours ago, ZooNamedGames said:

Well, to be realistic- the LM never docked with the CSM during Apollo- the CSM did.

That doesn't mean the LM couldn't only that they didn't.  In fact, the LM was quite capable of taking the active role during docking if need be.

The CSM was the active partner for a nominal docking because CMP was specifically trained to dock to the LM and had a better FOV.

 

5 hours ago, sevenperforce said:

Also on the subject of realism...EVA in orbit was...not really a thing for the LM.

As with docking, that they didn't doesn't mean the couldn't.  It was the backup method for transferring from the LM to the CSM.  (And they performed a stand-in-the-hatch EVA on orbit during Apollo 9 and on the surface during Apollo 15.)

[RoboRay]

The main problem with the MEM RCS is that the part's Center of Mass is at the base, rather than up in between the RCS thruster quads where it is supposed to be.  

This is with only a Docking Port Jr. and Spark engine attached, which seems to be the intended typical usage for ascending from the Mun.

A secondary problem is that the thruster quads themselves are attached asymmetrically, on both the vertical and longitudinal axes... which makes it effectively impossible to perfectly balance the craft.  Not actually impossible, but requiring far more patience than I posses.

I did some experimenting with offsetting the MEM part center of mass and found a reasonable compromise with torque around all three translation axes in the hundredths of a kN.

The values I settled on are:  CoMOffset = 0, 0.6998, 0.029  (the order of axes being lateral, vertical, then longitudinal)

You can download here a ModuleManager script that applies these values to (mostly) fix the MEM RCS balance.

[Commander Zoom]

Would making the MEM larger help with any of these issues?  'Cause I've thought since this DLC went live, and finally came here to post (and see if anyone else had), that it's just plain out of scale next to the CSM.

[EatVacuum]

It is a great looking model, although as mentioned it is proportionately smaller compared to the Mark 1-3 capsule than the real world LEM is to the Apollo capsule. It also apparently is only meant to be married to the junior docking port, the full size clampotron just doesn't work visually, totally ruining the otherwise great esthetics of the model. But while the Clampotron junior looks OK on the MEM, it really doesn't work well on the Mark-1-3 capsule, only the full sized docking port works there, although that can be addressed with an intermediate piece - stacking a small clampotron on top of the full size one works, and allows for the capsule to dock with both the MEM and (after detaching the junior) any station you have that uses regular size ports.

But ignoring the esthetics, It does seem to have gotten at best a quick, or no pass at all through the old QA team,  it is borked in at least three ways that impact actual gameplay.

1) As already mentioned, the RCS is unbalanced and it needs a CoM fix - using the "horizontal" plane  (I,J,K,L or WASD key depending on your preference) translational thrusters produces a torque causing rotational spin in the opposed direction.  While it is not terrible when SAS is on, it continually fires the RCS to fix the rotation, wasting fuel and making your docking maneouvers a little trickier. It is horribly obvious with SAS off, so much so that I have to wonder if any of the devs actually flew the MEM while working on it, let alone passed it by any actual QA testers or beta testers. It's hard to believe that this problem would have escaped even a casual play through of a Munar landing with orbital maneouvering, let alone any actual docking maneouvers. Miving the CoM upwards will fix this.

2) Not specifically mentioned previously - the RCS is also not centred with the CoM in the horizontal plane, so if you fire the RCS thrusters for movement forward or aft (the H,N keys), you get rotational torque causing yaw. Adjusting the CoM so the RCS is balanced will move the CoM horizontally, which means it will be offset from the centre of thrust of any engine placed on the normal attachment point, which will cause torque using the engine. That would be worse than having the RCS unbalanced as it will cause torque and yaw or pitch rotation issues during powered ascents from the surface or when using the engine for orbital maneuvering. Using the move tool to offset the engine to get the thrust aligned is frustrating and fiddly, so you probably won't get it perfect. And then, assuming your descent stage is balanced itself, it in turn will have to be adjusted to get its center of thrust back under the CoM. The only good fix is for the devs to rebalance the thrusters.

3) Doing an EVA does send the MEM flipping as previously mentioned, but that is the case with some other manned parts, so it is hardly unique although it does seem to be worse than most. This is not a big deal as you don't have to EVA to transfer crew back and forth for a classic munar mission, but it is an inconvenience if you want to use it for any operations that do require EVAs.

If I were a KSC engineer, this part would not pass inspection, it would get sent back to Sean's Cannery for a rebuild. Looks like it's back to my old ugly squat lander with the Mk2 lander can for now.

[LunchIn10Seconds]

Just started having a go with KSP after a long break. I'm going to start by saying I am not a fan at all of the take2 rebranding, the new UEA is pants and Squad have been relegated behind take 2s new 'brand' Private Division.

Still I suppose if you sell your Sole you're going to end up getting battered.

Which makes me wonder how much of a hand they actually had in the DLC. This lem remake looks like a modeller outside Squad was given the task of making one without ever seeing any of the game's other assets.

As mentioned its too small for the standard docking port, but its also a pain to connect to much else. As soon as I add an engine I get the starting of a fairing but that's wrong as the lem was packed behind the csm in Apollo.

I have no idea how to get a ladder attached - again these are out of scale with the lem.

Trying to build a realistic lander stack leaves gaps everywhere ...

And this is before I've tried to fly the thing, which to be honest I probably wont both to do given all the problems that have been raised in this thread.

Nice model, 0 QA hope this is not the way things are going to go.

Edited April 24, 2018 by LunchIn10Seconds

[jamqdlaty]

CoM of this part is still not fixed. It sounds like a bad joke.

[Errol]

On 3/21/2018 at 2:31 PM, RoboRay said:

I did some experimenting with offsetting the MEM part center of mass and found a reasonable compromise with torque around all three translation axes in the hundredths of a kN.

The values I settled on are:  CoMOffset = 0, 0.6998, 0.029  (the order of axes being lateral, vertical, then longitudinal)

You can download here a ModuleManager script that applies these values to (mostly) fix the MEM RCS balance.

Since this link isn't working anymore and, as has been pointed out since, this is still borked in stock many years later, I decided to re-create said script using the suggested values myself. I also found that one of the bottom stacking nodes was interfering with the other, so I modified it as well, and now it is easy to attach parts to either of the two bottom stacking nodes, allowing you to choose if you want that hideous skirt adapter.

 

@PART[MEMLander]:FINAL
{
	%CoMOffset = 0, 0.6998, 0.029
	%node_stack_bottom2 = 0.0, -0.25, 0.0, 0.0, -1.0, 0.0, 1
}