[1.9.1+] OPT Legacy 3.1.2 | Reconfig 3.4 [Apr 20, 2021]

[JadeOfMaar]

On 10/11/2020 at 5:41 PM, LesnoyLis said:

First of all, let me express my gratitude about your work. Thank you for upkeeping that outstanding mod.

In second, excuse me for my bad english.

And now, let me ask a few questions.

Currently, CKAN shows only OPT Reconfig support 1.10, OPT Continued and OPT Legacy shows as 1.9.1 If I manually install both, some parts shown as "legacy" and some of them has dupes. Any way to solve this?

I'm happy that you're satisfied with my maintenance work. It means a lot. But I cannot fathom how and why manually installing the mods makes trouble for you. If anything, you should be less exposed to problems than if you let CKAN install them. Some advice to installing these mods, and mods in general:

• Always delete a mod's folder from GameData to prevent the chance of files that have been renamed or removed during an update, to continue to exist alongside the files that replace them.
• Know where, relative to GameData, that the mod folder inside the zip file is positioned. It is likely that you accidentally placed an OPT folder inside of another mod folder which is the most obvious cause for duplicate parts and repeating patches.
• In CKAN, go to Settings menu -> Compatible KSP versions -> then tick the tick boxes for other KSP versions. In this situation you only want 1.9 and 1.10. Click Accept. Your OPT/CKAN troubles are now gone.

On 10/11/2020 at 8:17 PM, Nicky21 said:

I don't know if you fixed this in the latest release but the OPT ACS part crashes the game like crazy. I am talking about the single port, not the one with 4 variants. It even crashed my game once when i clicked on it in the vab. I think it's sound related ??? I'm sorry, I can't provide a log.

ksp 1.10.

Remove any mod that adds special effects to RCS ports. I assume somehow those mods are conflicting with B9 Part Switch. As far as I know, only mods that I've contributed to have configs for RCS fuel switching so this may be quite an edge case. My mod installs are very light and narrow of scope so I'll never experience for myself most of these conflicts and issues that people come up with.

I assume you mean the "linear" one, not the retro, and definitely not the aerodynamic OMS. I'm not aware of an OPT RCS port with four variants. The ones that can change colors only have 2 options: gray; black.

Edited October 14, 2020 by JadeOfMaar

[LesnoyLis]

1 hour ago, JadeOfMaar said:

I'm happy that you're satisfied with my maintenance work. It means a lot. But I cannot fathom how and why manually installing the mods makes trouble for you. If anything, you should be less exposed to problems than if you let CKAN install them. Some advice to installing these mods, and mods in general:

• Always delete a mod's folder from GameData to prevent the chance of files that have been renamed or removed during an update, to continue to exist alongside the files that replace them.
• Know where, relative to GameData, that the mod folder inside the zip file is positioned. It is likely that you accidentally placed an OPT folder inside of another mod folder which is the most obvious cause for duplicate parts and repeating patches.
• In CKAN, go to Settings menu -> Compatible KSP versions -> then tick the tick boxes for other KSP versions. In this situation you only want 1.9 and 1.10. Click Accept. Your OPT/CKAN troubles are now gone.

Thanks a lot for your answer! Problem are solved =)

[AmpCat]

@JadeOfMaar, looking at your mods (picking up KSP again after a bit) and love the improvements. I was playing with the Egg Dog engines, but I have an issue with the variable ISP feature. Where it should have a variant box to select the ISP like in the screenshots, I just get a text field that says 'VIsp Step: 540'. Same on both models of Egg Dog. Thoughts? Or do you need logs and/or other details?

[JadeOfMaar]

@AmpCat I assume you're in a progression game. If so, the feature is partially locked by the part upgrade system. Look for these and other circular icons in your tech tree.

Edited October 25, 2020 by JadeOfMaar

[AmpCat]

Yep! Makes sense now. I saw those a little further down. Thanks!

[Daniel Prates]

Guys, how to you manage to keep your SP oriented during reentry? I have tried several designs and  they all eventually tumble like falling rocks, I only regain control when they have already slowed down around 20k meters or lower. 

 

[JadeOfMaar]

@Daniel Prates Some basic and not-so-basic things to help you out:

1. Ensure that your CoL is very close to your CoM and you have plenty of elevon area. The further from the CoL, the more they're effective.
2. Have your wings high mounted so your CoL is above your CoM. This will add paraglider/parachute behavior to your plane and produce a lot of extra stability when you need it. This is the opposite effect to the waverider-- it wants to fly upside down problem of low-mounted wings. (Sadly, IRL supersonic and hypersonic fluid dynamics would be very hostile to this, therefore, hypersonic craft designs for IRL seem to always be waveriders.)
3. Add extra elevons or large stabilizers and (depending on your spaceplane's intentions) consider disabling their authorities, reserving them for deploying and adding a static deflection force to help you stay stable. Even if you do well with point #1, you may still end up with too much unwanted deflection to cancel out. Space Shuttle has huge elevons and that huge center flap after all. Civilian airplanes and fighter planes have extra elevons (flaps, slats, spoilers...) that deploy for lengths at a time: just to add a constant pitch momentum; lower their stall speed; or break their lift potential after touchdown so the craft doesn't accidentally lift off again.
4. If you have tall vertical stabilizers, consider putting OPT airbrakes up there to serve as Space Shuttle's split tailfin airbrake. Airbrakes in KSP can be used very well to create pitch and yaw moments once you place them well and you know that they'll respond to pitch and yaw commands just like elevons.

This spaceplane (available on KerbalX), can (oddly enough) easily hold extreme AoA during reentry. Its wings are high and dry to not cause thrust torque by holding fuel, and to add that paraglider behavior so that it's tumble-proof.

[Daniel Prates]

Thanks @JadeOfMaar! 

[caipi]

I'm just gonna leave this message here in case anybody else runs into the same issue, so they don't have to chase down the error for an hour or more. There is no further need for assistance and this was completely my fault!

My issue was, that several OPT-parts, such as cockpits etc., had multiple (two, to be exact) ModuleCryoTank nodes. This paired with Dynamic Battery Storage (part of Nertea's Cryogenic Engines) created NREs en masse:

"Multiple ModuleCryoTank nodes found with identical or no moduleName"

This was caused when I updated from OPT Reconfig 2.1 (or any previous version) to 2.2. I foolishly made the mistake to just unzip the new files into the GameData directory without deleting the old OPT Reconfig folder first. I did it, oh well, because I was lazy. With Reconfig 2.2, the OPT_Reconfig\CRP\OPT_B9PS_Cryo.cfg changed to OPT_Reconfig\CRP\OPT_B9PS_CryoTanks.cfg, so now two similar patches existed which did similar things, adding the CryoTanks modules twice. -.-'

Lesson learned. Always make a clean install with new patches, no matter how insignificant the patch seems and how excited one seems and how annoying it is to navigate to the game folder to delete the old folder. I'm sure I'll remember it. At least for a month or so. I promise. Well, okay, let's say at least two weeks. I mean, it's not like I didn't know it before...

[Daniel Prates]

@JadeOfMaar perhaps you can further help me here. 

This is  my initial design (which didn't work), prior to me asking your help. With two thousand hours of KSP I thought I have refined everything, but clearly this may be a good design for a rocketplane, but not for atmospheric reentry:

Flies like a charm after takeoff, but the thing falls flat when the issue is reentry. Clearly a design good in general for overall flight, does not necessarily do reentry well.

Then, as per your suggestions, I redesigned the thing, as the pictures below show. Higher wings, upper CoL, CoL much closer to the CoM, more pitching surfaces, even the rudders double as yaw+pitch surfaces. And I moved around the airbrakes, they now form a sort of clamshell design, with surfaces above and below.

 

Turns out that doesn't help too (though it is much, much better now, or less bad anyways). The thing starts well but eventually yaws and pitches out of control. It just seems to have a mind of its own to fly bottom-first (giggity):

 

 

 

Perhaps I am lacking more yaw control?

The airbrakes are perhaps to blame, aren't doing what I assumed they would do? I imagined they would add drag at the rear and help stabilizing, but then again, maybe that doesn't matter (as per the 'pendulum falacy'). 

Did I add insufficient pitch control, when I actually need MOOOAAARRRR instead of just 'more'?

How about yaw control, do I  need cartoonishly large rudders?

How would you fix that design?

Edited October 28, 2020 by Daniel Prates

[JadeOfMaar]

@Daniel Prates Hmm.

1. I tend to avoid placing angled wings on spaceplanes. I personally don't do well with them. And I think no hypersonic vehicle concept has them except maybe for the tailfins. You will always have some amount of non-zero roll amount thanks to wobbly stock SAS or no autopilot mod, so your lift will never truly be centered on and supported by the lifting body (the spaceplane's fuselage) and in thin atmosphere it's easy for one angled main wing to stall out and cause the plane to start to roll wildly, which is probably what you're experiencing.
2. Whenever I do place angled wings or control surfaces, I always disable yaw because the line gets blurred between yaw and roll. I disable yaw unless I put angled wings the opposite way (some swooping down, anhedral, to match and balance the ones swooping up, dihedral). Angled wings add roll instability which is very welcome for stunt planes, fighter planes and airliners. (For crewed aircraft, especially airliners, it's implicitly forbidden and it's just a very bad idea to use actual yaw. See: Coordinated Flight/Turns where you replace yaw with a procedure that involves roll then pitch).
3. Rotate your plane by its root part in the SPH. The CoL will move based on your AoA. If you rotate far and the CoL goes or stays in from of the CoM, the design needs to change.

Where did you get those airbrakes from? I'm intrigued. I also wonder if they even work well granted their size.

 

[JonnyOThan]

2 hours ago, Daniel Prates said:

How would you fix that design?

I'm not great at planes, but the one thing that sticks out to me here is that your CoM is closer to the tail than the nose.  This is the opposite of what you want for stable flight.  Ideally it would be near the center of your craft, but if it's closer to the nose then it will be more passively stable.  The reason is that even though you have draggier parts at the tail, the torque caused by drag is the force of the drag multiplied by the distance to the center of mass (more specifically, the distance to the axis of rotation).  So a part that is half as draggy but twice as far away will torque the plane just as much.  I suspect something around the nose of the plane (or maybe the canards) is creating enough drag to flip you around.  Similarly, since the airbrakes aren't very far behind the CoM then they don't do much to stabilize the ship.

Edited October 28, 2020 by JonnyOThan

[JadeOfMaar]

@JonnyOThan You have a great point there. In my personal experience, spaceplane cockpits tend to be quite draggy (at least as much as the cabin or longest fuselage piece) and they can easily bring the CoP forward. The nose canards, being anhedral, will try to produce yaw (but actually produce roll) when they shouldn't, in addition to the issue of cockpit drag, which can make the plane much more unstable, but I've already covered that in my answer above.

[Daniel Prates]

Great conversation @JadeOfMaar and @JonnyOThan. I will do some experimenting and report back.

Obviously SPs and proper airplanes are different animals altogether.

Hmm... let me drop this by real quick:

1. reactions wheels. You guys disable them? Different control inputs counteracting eachother is a know cause of weird behavior;

2. You guys keep SAS active, for sure. Does it suffice or constant manual inputs are also needed?

[JadeOfMaar]

@Daniel Prates Yeah, they definitely are very different breeds, needing very different care.

1. Reaction wheels in cockpits? I know a few players who go out of their way to strip those out, and I've made configs to help them. Dedicated reaction wheels parts (drone cores and inline SAS) tend to suffice for RW power.  In most cases outside of atmosphere you don't need much of it. In the cases of needing to turn very fast/forcefully, like for an emergency condition while landing in vacuum, or resisting flipping out on reentry, always be sure to have plenty RCS fuel and RCS thrust. Meanwhile, too much RW can become a problem (especially for long, noodle-ish spacecraft). In atmosphere it can be argued that you don't want to lean on RW torque (for several good reasons I can think of) and youdl rather have air-breathing RCS (which is actually hard to come by, but OPT is one mod that has them).
2. I tend to keep SAS on. But it's arguably better to have an autopilot mod. Stock SAS is actually pretty miserable. It easily misjudges and overcompensates, and it gets worse and worse if you have engines with slow gimbal response and/or wide gimbal range, and if your aircraft is prone to rolling like your spaceplane with dihedral wings. An autopilot will hold your pitch, roll, heading, cruise altitude, cruise speed, all sorts of cool and precise things like that for long or tedious flights. Typically, an autopilot mod will disable SAS, as SAS's logic will conflict with it. There will always be a need for manual input during the flight of any craft in KSP. But it depends on the player how well he or she can build, and how much he or she enjoys manual controls.
   
   In my case, especially, I rarely play, and often tend to just do test-flights of spaceplanes so I'm lazy to install an autopilot and I'm pretty confident with manual controls. On the other hand, Jonny is a kOS user so a lot of what goes on during his flights (which may not involve spaceplanes at all. It's been a while since I've seen his Twitch streams) may be controlled by effectively doing MS-DOS commands. SAS would very likely be enabled between the executions of kOS commands.

[AmpCat]

@Daniel Prates There's also a mod that does a better CoL and calculates your yaw pitch and roll stability at different altitudes and velocities. Your speed has a big impact on the effectiveness of control surfaces and stability in general. I've had the same troubles, and using that mod has eliminated them for me. The first one is called 'CorrectCoL', I've not used it yet in 1.10, but in the past it's worked well. It doesn't take into account velocities, but it does a better job taking into account angled control surfaces and canards. Basically, in reality, your CoL is probabaly not actually behind your CoM, despite the indicator in the hangar; it's just not right. This should fix it. But it takes a lot of computational power.

CoL also gives you a new button that looks like little graphs in the vehicle editor. It shows you these properties where you can specify a range of altitudes, AoAs, etc. and then plots the stability of that axis.

Edited October 29, 2020 by AmpCat

[gilflo]

On 10/26/2020 at 5:59 PM, Daniel Prates said:

Guys, how to you manage to keep your SP oriented during reentry? I have tried several designs and  they all eventually tumble like falling rocks, I only regain control when they have already slowed down around 20k meters or lower. 

 

Hello

My golden rules to built SSTO

-No angled wings, no angled fins, no angled tails

-The COG with and without fuel must be very near ( maximum 1 yellow or red circle)

-Always use RCS Build aids to minimize torque, it can take a long time to do this but it's very helpfull. And avoid Nose or Tail pitch

-Center of lift must be at least "one circle" or more behind COG because as you accelerate it moves. That can give you less maniability or control in low altitude, but you can mitigate using Speedbrake for pitch in low atmosphere if you need . "Circle"s are the red or yellow circles of COG

-Big tails and numerous tails are greatly helpful for yaw control, the bigger the plane the taller the tails.

-Use CorrectCol mod to check your Pitch and Yaw are stable in the whole flight enveloppe, from ground to orbit

 -Atmospheric Autopilot which is a manual fly by wire mod can be helpful with big SSTO

Here is a small SSTO for Eve ( you will understand what I am talking about with "circles": Here the COL is more than "2 circles" behind COG and I use speed brakes to help for pitch control in low altitudes and speed

See that at 60000m and 2000m/s it's still stable on the COL graph

Here is a much bigger Hydro SSTO. Notice the 6 big tails and the numerous elevons at the rear to help for pitch - "Canard or Moustaches" are also helpful when the SSTO is long to help for pitch torque.

For re-entry, use retrograde, cockpit facing earth, Roll 0°, then once you enter atmosphere, use the "cobra" position with "Radial out" on the Navalball.  Vernor engines on and under the cockpit for pitch attitude and on the wingtip for yaw can be helpfull to maintain this position ( belly facing heat) until under 1500m/s

The SSTO is a big speed brake in this position. Then select Prograde and when you are stable, select Surf or disconnect and pilot.

Edited October 29, 2020 by gilflo

[DeadJohn]

On 10/28/2020 at 6:27 PM, Daniel Prates said:

1. reactions wheels. You guys disable them? Different control inputs counteracting eachother is a know cause of weird behavior;

2. You guys keep SAS active, for sure. Does it suffice or constant manual inputs are also needed?

My comments are focusing on deorbit because that can be the fussiest part of SSTO flight. Ascent is more forgiving.

1. I leave reaction wheels active. It means less RCS fuel burnt for docking maneuvers, and most reentries can be done without RCS. Reaction wheels and control surfaces usually cooperate with each other.
2. SAS often needs manual corrections. My usual spaceplane deorbit has me coming in around 25 degrees nose-high to bleed off speed. Some planes can hold that pretty well, others tend to want to pitch up. I use Stability Assist SAS, and either tap 'w' to occasionally lower the nose, or switch to Prograde Hold SAS then quickly back to Stability Assist before the nose gets too low.

Most of my spaceplanes utilize some form of canard near the nose. That provides strong pitch authority, extra lift, and makes takeoffs simple. The only downside to canards IMO is increased risk of flipping if I get too nose-high during a fast reentry; they grab the air and cause the nose to have more drag than the tail.

This screenshot shows an example of canards and reentry. Note that the twin rudders look angled but they are perfectly vertical; as others already mentioned winglike surfaces are best vertical or horizontal to avoid unwanted control interactions.

[gilflo]

Here is an example how to deal with atmospheric re-entry using Vernor engines on and below the cockpit: It's at the beginning of the vid

 

[IronMaiden]

On 10/28/2020 at 10:58 AM, Daniel Prates said:

This is  my initial design (which didn't work), prior to me asking your help. With two thousand hours of KSP I thought I have refined everything, but clearly this may be a good design for a rocketplane, but not for atmospheric reentry:

Flies like a charm after takeoff, but the thing falls flat when the issue is reentry. Clearly a design good in general for overall flight, does not necessarily do reentry well.

Then, as per your suggestions, I redesigned the thing, as the pictures below show. Higher wings, upper CoL, CoL much closer to the CoM, more pitching surfaces, even the rudders double as yaw+pitch surfaces. And I moved around the airbrakes, they now form a sort of clamshell design, with surfaces above and below.

Turns out that doesn't help too (though it is much, much better now, or less bad anyways). The thing starts well but eventually yaws and pitches out of control. It just seems to have a mind of its own to fly bottom-first (giggity):

Perhaps I am lacking more yaw control?

The airbrakes are perhaps to blame, aren't doing what I assumed they would do? I imagined they would add drag at the rear and help stabilizing, but then again, maybe that doesn't matter (as per the 'pendulum falacy'). 

Did I add insufficient pitch control, when I actually need MOOOAAARRRR instead of just 'more'?

How about yaw control, do I  need cartoonishly large rudders?

How would you fix that design?

Your problem here is as Jonny0than mentioned, your CoM is too far back. That's an inherent problem with putting engines on the back of your plane. It gives your cockpit way too long of a lever arm to torque your plane around and it gets even worse as you drain fuel. The problem is hidden by the fact that the game only used wings and control surfaces in the calculation of the CoL. Mk2 parts are considered wings as far CoL is concerned but they create more drag than regular wings and that's not accounted for. It is a center of "lift" after all, not a center of "pressure" which is what we're really after. The best way to accommodate for this is to put your CoM in the middle of your plane so that your fuselage is neutrally stable. Here's a little guide which goes more in depth on this issue.

There is a way to help the design a bit if you absolutely must keep the engine in the back aesthetic. That is to move your main wings further back and use wing incidence to move the CoL forward at low angles of attack, and backwards at high angles of attack. Wings further forward should have higher angles of incidence and the angle should get closer to 0° as you move back. Now at low angles of attack CoL moves forward for efficient and balanced flight, but if you try to pitch up or down too much the CoL will move backwards to put a stop to runaway pitch. It is self correcting and the plane will tend towards a few degrees of prograde. This is not as efficient as having a properly balanced plane but it's not bad for hamfisting some stability on to an aesthetically pleasing but not necessarily aerodynamically pleasing design

[DeadJohn]

On 10/28/2020 at 1:58 PM, Daniel Prates said:

SNIP... Turns out that doesn't help too (though it is much, much better now, or less bad anyways). The thing starts well but eventually yaws and pitches out of control. It just seems to have a mind of its own to fly bottom-first (giggity): ... SNIP

I just took another look at your plane and have some additional ideas.

If a plane reaches orbit okay but flips to easily during reentry, you might be able to remedy that just by shifting fuel between tanks. Move some fuel from the bicoupler to the forward tank when you're leaving orbit. If you move too much and find your plane's nose drooping in the low atmosphere, you might need to move some back.

Dumping some fuel evenly from all tanks (I still use the Hyperedit fuel slider rather than the new-ish fuel valves) before deorbit can also help; less mass makes the aero surfaces, RCS thrusters, and reaction wheels seem more powerful.

I can't tell if the middle bay of your shuttle is a fuel tank or cargo. Cargo shuttles add more difficulty. Even if you put the load right at the center of mass, planes can fly very differently with different masses.

 

 

 

 

 

 

[Daniel Prates]

Right, well, to  begin with, my many thanks to @JadeOfMaar @DeadJohn @IronMaiden @gilflo @AmpCat and @JonnyOThan. Following your advice I think I have honed my skills in SP design by a great degree.

This is what I did with all your input:

It can take more refinement but condenses all we spoke about the previous days: It has the CoM much closer to the CoL when near-empty, very angular surfaces, balanced drag between nose and tail due to an overall 'diamond' shape of sorts, few but powerful control surfaces, no clutter  of control surfaces that could cause minute lift or drag inputs where I don't need it. It looks bad but did the trick: my first SSTO take-off and landing without loss of control:

I overshot KSC by several dozen kms but, with those wings and near-empty, gliding back was no sweat:

Many thanks to you all, it is this kind of discussion that makes KSP so great.

What next? I think I will angle the wings back to get better pitch control, raise the wings so that the CoL is higher, and think about another engine setup or combination. My concept here is kinda flawed. I have only upper-stage rocket engines, so it saves weight and does well on higher altitudes but suffers during most of the climb. It barely has enough Dv to reach a LKO and do something useful up there, but hey, it's a start. Nice sunday y'all!

Edited November 2, 2020 by Daniel Prates

[Daniel Prates]

Not to drag my write-up for too long, I'll just post here the final refinement: wings swept farther to the back; moved them up so the CoM is below the CoL; CoM wet and dry were brought closer together (they almost coincide), and both are very close to the CoL.

No surprise here: it handles very, very well now:

[IronMaiden]

3 hours ago, Daniel Prates said:

Not to drag my write-up for too long, I'll just post here the final refinement: wings swept farther to the back; moved them up so the CoM is below the CoL; CoM wet and dry were brought closer together (they almost coincide), and both are very close to the CoL.

No surprise here: it handles very, very well now

Vastly improved weight distribution dude, well done!

[Daniel Prates]

I used reaction wheels during the initial descent, then the fky-by-wire feature of AtmosphericAutopilot for the more grinding, heavy deaccelerating portion.  The surfaces were all over the place, the rudder specially, but with such excess of control area, it managed very well.