Test of Russian manual docking system doesn't quite go as expected
Russian ISS docking system test doesn’t go as planned, SpaceRef (With video)
“According to veteran Russian space program reporter Anatoly Zak an ISS test of the cosmonaut-operated docking system on the Progress 62 cargo spacecraft didn’t quite go as expected earlier this morning.”
Marc’s note: Despite a statement from Roscosmos saying the test was successful you can watch the video yourself and see docking with considerable pitch at the end. And yes, there’s a reason these tests are performed and I’m sure there will be another scheduled not in the too distant future.
#Roscosmos #NASA claim success in Progress docking test: https://t.co/kty9zKcjIJ So don't believe your eyes and ears pic.twitter.com/R7zTUHxegg
— Anatoly Zak (@RussianSpaceWeb) July 1, 2016
Gee, remember back in the 1990s when the Russians “tested” the TORU manual guidance for a Progress coming back into Mir, and it crashed into the SPEKTR module and Mike Foale et al had to cut the umbilicals between modules with hatchets and close the door… and then we found out that they were testing the TORU because the cosmonauts got paid more for a manual approach and docking than the regular KURS method.
Hmmm… Good times.
(smile)
jamesmuncy Yes and all of Mike’s personal stuff and much of his experiment hardware was in Spektr.
The nice cosmonauts did an IVA into the airless module and retrieved some stuff – including his toothpaste!
Actually the reason for that manual test was different. Due to the breakup of the Soviet Union, they wanted to do away with some of the automated docking electronics which was not manufactured in Russia, but in Ukraine.
http://spaceflight.nasa.gov…
That was one lousy docking. Rotation rates (peak) of ~1.8 degrees per second are, shall we say, significant (the rotation rates are the WX, WY, & WZ (W=omega) in the upper right of the docking camera view; orbital rate—one spin around the Earth ~360 deg/90ish minutes—is about 0.067 deg/sec).
And, NASA PAO’s gift for going to the WRONG camera AND talking inanities over the action (about ‘the great television view!!!) at JUST the critical moment comes to the fore. Arrggghhh. The Russians are known for flying things hot (should I say ‘authoritatively’?), but the behavior (at least superficially) following the big final ‘shove’ to initiate final approach and docking smells to me like an orbital dynamics cross-coupling, whereby a shove in toward the station may have caused a lateral shift (can’t tell for sure since they took the best views away at just the right time). If so, such lateral ‘drift’ might have caused the Progress probe to strike outboard on the ISS’s drogue cone.
With the towards-the-ISS shove he obviously gave the Progress, it had enough forward momentum to continue down the cone slope into the receptacle, producing a capture. [Remember the scene in Apollo 13 where Kevin Bacon is potentially frying in the pan?) However, the same slide, pushing the tip of the Progress laterally, might have induced the high rotation rate (mostly in the Z axis, which would translate (generically at least) in an American system to a yaw…not sure how the Progress vehicle was oriented relative to the LVLH frame & the ISS frame), which is evident in the docking camera view and then the external view. Somewhere in the docking sequence the vehicles (at least the Progress, I would think) would switch their attitude control system to Free Drift, which (if it happened early enough), would have permitted such an attitude deviation for the Progress. Ouch.
Lots of speculative fill-in for sure, but that’s my best guesstimate with no further info to go on. It could just as easily have resulted from a last minute lateral correction (or over-correction)…or a computer or control stick glitch. Hard to say with anything approaching confidence.
Saw this this morning:
http://www.spaceflightinsid…
Says that normally there is an automatic in-line firing at the moment of first contact to force the engagement of the latches (this matches fairly closely how the shuttle had docked using the Russian-designed APAS) but they got a lateral one as well.
In 1960-1962 as plans were being made for a moon landing there was great discussion about the “mode” (Direct-ascent, Earth-Orbit Rendezvous, Lunar-Orbit Rendezvous, and Lunar Surface Rendezvous); while the discussion focused on the size of the initial rocket leaving earth, and on the rockets needed to leave the moon, the ability to actually rendezvous or dock two spacecraft in space was a big part of the discussion.
All of the SF before supposed that we’d easily move from one spacecraft to another. Some imagined mothership hangers, others the sort of very primitive docking currently state of the art. But none imagined just how difficult it would be, requiring either a judiciously-planned launch or an amazing amount of fuel.
And a fair amount of skill, apparently.
After several decades of mostly event free docking we probably do take it for granted. On the surface it seems similar to landing an airplane. Planes may have wider margins, but land at high speed. Docking has very tight margins and proximity within inches of another vehicle, but it’s done at very slow speed. But the risk is always there. A thruster failure at an inopportune moment. A control not responding just when you really need it. Capture mechanism that doesn’t function correctly. Not to mention errors by the pilot, whether human or automated.
Maybe in the future this type of docking will in fact be seen as old fashioned and berthing will be the norm. Unless there is some type of advantage to the type of capture mechanisms that require momentum to achieve initial soft dock.
And if they have automated docking there is no reason not to have automated berthing, a computer operated robotic arm should have no problem grappling and berthing a spacecraft.
Docking is in the hands of the spacecraft doing the docking; berthing requires an active intelligent system on the object being berthed to. No functional arm, no chance of berthing.
But if the approaching spacecraft can be flown actively, a link-up is still possible even with a completely passive ‘target’ spacecraft.
This is the philosophical rationale behind maintaining a docking capability.
Initial grapple by the target-based ‘arm’ (or whatever device) has to manage, very delicately, the same free-flying dynamics as docking. While berthing puts less burden on required capabilities of the approaching spacecraft, grappling & berthing involves similar challenges.
I know it’s not a perfect analogy but I have one of those miniature drones that you can fly indoors. Its home “base” is on my bookshelf. I can either fly it and land it onto the bookshelf, or I can pick it up with my hand and place it onto the bookshelf. Even though essentially the dynamics of what I am doing are the same, the first method is much more difficult and has quite a risk of mission failure, no matter how good I get at flying it or how many times I practice. However using the latter method it is quite easy to safely “berth” the drone onto the shelf. So it’s just hard to imagine that berthing with a robotic arm is as challenging and dangerous as docking.
Although I realize that one (of many) oversimplifications of my analogy is that when I “grapple” the drone with my hand it is sitting perfectly still on the table, whereas a spacecraft being grappled is free floating and there could be some slight movement between the two vehicles. Perhaps that is what you are referring to. But assuming this movement is minimal then I would think the human or computer should have no problem slowly and leisurely moving the grapple device into position over the pin. However if there is a lot of movement between the vehicles which requires fast reaction times by the arm then I can see where it could be challenging.
Understood that if it’s difficult or impractical to place a robotic arm on either vehicle then docking would be required. But the only scenario where I can imagine that being the case is when both vehicles are passing through an atmosphere, such as a transfer vehicle carrying a crew from the surface of Mars to Mars orbit where it docks with a vehicle that will renter through Earth’s atmosphere. But as long as at least one of the vehicles will be spending its working life in space it can in theory have a robotic arm attached for berthing. Realize there is size and mass involved for the arm, and if there are multiple docking ports this means multiple robotic arms. But this is an area of tremendous progress so I just envision robotics becoming the norm for autonomous docking of two pieces of anything together in space.
I have some experience in this subject matter. Your analogy IS way off because the dynamics you are leaving out make all the difference. Float your drone on an air hockey table for your ‘simple’ arm grab and you’re beginning to get closer to a reasonable analogy. [If you’re sitting & floating on the table, too, even better.]
Once grabbed, I freely acknowledge, the process of making a union between interfaces is potentially more controlled (while noting that using mechanical articulation introduces its own issues that must be addressed). But before grapple, approach & docking shares many aspects, dynamically speaking, with robotic capture. While a docking pilot doesn’t have to deal with arm wobble, he must contend with attitude control dead-banding. Bump the spacecraft either way, and you’ll see the similarities fast.
But I think you missed my main point, one that hits (ha!) on fundamental philosophies of spacecraft hardware & operations design that in fact were thoroughly discussed for Freedom & ISS. Suppose a Cygnus launches to ISS and the SRMS doesn’t work: no cargo delivery. Ditto Dragon & HTV. Not so Progress or the ATV.
If berthing is the only scheme for ALL deliveries to ISS, you will have designed in a potential failure mode that cripples your entire endeavor. Think further: what if, sometime in the future, the same berthing-only station needs to be evacuated for a major problem that similarly makes grapple of any returning spacecraft (cargo or crew) impossible. Again, you have designed into your systems & procedures a potentially program-halting scheme.
Docking, by ‘formal’ definition in space ops land, means that the approaching spacecraft maintains control over linking up. Such a system avoids the two scenarios I outlined above and offers redundancy in both hardware & operational modes.
Both ways have their advantages & disadvantages. Choosing one completely over the other cuts down operational flexibility.
Well I tried to make it as clear as possible that the drone analogy was limited, thanks for voicing your agreement with me on that! But seriously, you did get my main point which is that after grappling has occurred it should be easier to connect the two spacecraft.
As for difficulty grappling I would think it’s all about rates. In a perfect setting, like I acknowledged was a limitation of the drone analogy, the two objects are stationary. But in theory even in space they can be stationary relative to each other, allowing a more leisurely and careful positioning of the end effector over the pin (avoiding contact of course). Understood that some drifting and roll on the three axes will be occurring, how much this can be limited to I don’t know. But since the docking spacecraft, after reaching acceptable rates will presumably be placed into free drift, the rates will be steady and predictable, unlike the constant change in rates for a docking spacecraft with thrusters firing. With steady rates I would think a computer (or human with lots of skill) can adjust for it and also determine before making an attempt if the grappling is possible with that rate. And aborting a grapple at the last second seems like it would be easier than aborting docking in the last few seconds. Now if the rates are high and for whatever reason cannot be lowered, then I can see that grappling could be as difficult as docking, in theory even more difficult.
To push my luck with one more imperfect analogy, in golf it’s easier to chip onto the green than to chip the ball directly into the cup. So I would assume that it’s also (relatively) easier and safer to position a spacecraft 10 meters away from a another spacecraft than hitting a docking target. Again not easy, and not perfectly safe, but I would think relatively easier and safer.
As for failures, a broken docking mechanism puts the port out of commission and is not easily fixed. But probably that is less likely than a robotics hardware failure. But arms can have redundancy, especially the type of arms that can move around the station, in which case they just need to make sure there is more than one arm attachment point near the port in case a failed arm is stuck on one of them. But maybe we’re not there yet in terms of reliability for the robotics. I certainly don’t question your knowledge on the subject. But again I was referring to where it may eventually evolve.
But in theory even in space they can be stationary relative to each other, allowing a more leisurely and careful positioning of the end effector over the pin (avoiding contact of course).
You are neglecting the effects of naturally imparted relative motion caused by orbital dynamics. Even with distances under 50 feet such drifting occurs, complicating a grapple, potentially more so than a docking, especially when it’s time for metal to meet metal at first contact/touch.
As for potential difficulties during those last few seconds, watch the video of the STS-49 Intelsat capture attempts #1 & #2. Dumb scheme prone to failure, but what happened demonstrated that a free-floating object is still a free-floating object with which few folks have common experience. Imagine, with your original analogy, instead grabbing the drone out of the air with your fingertips…but without its rotors able to keep it stable in attitude or position for you…oh, and YOU are floating in the air, too.
As for failures, a broken docking mechanism puts the port out of commission and is not easily fixed. But probably that is less likely than a robotics hardware failure.
You are zeroing in on a portion of the rationale behind my primary point. Remember, I am talking about ensuring core/baseline functionality of the entire facility, not just the common need (across both methods) to make a seal between two hatchways. Docking capture mechanisms by design are made to be mechanically simple and therefore limited in possible failure modes (and, therefore, relatively easy to repair as well).
Docking itself, by design/definition, gives the approaching spacecraft the capability to affect a link-up regardless of the active/passive status of the space station. Berthing, on the other hand, requires a functional (& intelligent) RMS on the station, active station attitude control, PLUS all the same control requirements for the approaching spacecraft as docking requires (e.g., both berthing & docking s/c must maintain attitude & translational control of themselves to affect a grapple/docking while avoiding a collision).
Berthing introduces a suite of failure modes that risks the basic utility of the entire endeavor, failure modes above and beyond the failure modes that both methods have in common. Yes, it offers some advantages but it also introduces these additional risks stemming from the complexity of its absolutely necessary components.
I’m fine with accepting risk by using berthing in future space operations since it provides certain benefits (e.g., a less robust structural design of a cargo vehicle), but I would be VERY reluctant to eliminate docking completely when the entire mission depended on successful link-up.
Watching a video on the Aviation Weekly website showing how new F35s were refueled as they flew across the Atlantic to Great Britain I wondered if the Navy’s refueling technique- the use of a drogue – couldn’t be adapted to spacecraft. Certainly the relative speeds are about the same, after all, and finding a large-ish drogue is much easier than aligning the two spacecraft on the centerline.
That’s essentially what the Russian probe and drogue docking system (used on Soyuz and Progress) does. APAS docking systems (there have been revisions over the years) replace the probe and drogue with androgynous petals which serve much the same purpose.
Actually, many people imagined how difficult it would be, even OVER-imagined, which is one of the big reasons Direct Ascent held on so long as a contender. But Gemini as a program confirmed that it wasn’t as difficult as some had imagined. And since then, successful spacecraft docking has served as a functional component of many dozens (hundreds?) of space missions.
As for ‘primitive docking’ being current state of the art, I think this is a bit overblown. Granted, it’s not tractor beams & transporters. But one could just as well claim that parking one’s car in a garage or pulling a boat into a slip is primitive; they, too, each require a fair amount of skill.
The primary challenges of all these tasks involve the movement of sizable masses inside the domain of Newtonian physics. And yet lots of folks do such things every day.
I take your point that ‘primitive’ is an overstatement, not meaning to decrement the skills of astronauts. But I’ve wondered as well why manual docking is even practiced, aside from perhaps a ‘backup’ in the event computers fail. Shouldn’t computers do a much better job?
In a way, I’m reminded of ‘spam in a can’, the argument being that crew should actually do something. But isn’t this something that can be done better by computers?
Things are certainly drifting that way.
While this certainly could be completely automated, when you provide an astronaut with a manual override, the temptation is to take over at the slightest sign of trouble (real or imagined). This is doubly true if you pay the astronaut a bonus for manual docking (which I’ve heard the Russians do).
That said, have you ever watched a dock worker in an overhead crane loading and/or unloading a cargo container ship? They’re extremely proficient at what they do, which is maneuvering very large, very heavy shipping containers on top of each other with speed and precision that could make an astronaut jealous.
Actually I HAVE watched at the Port of Long Beach. Thanks for reminding me. Those guys are poetry in motion, to use an old phrase; I’ve seen them back-swing huge loads, sweeping a crane with simultaneous translation, only to have the load come to rest perfectly mere inches from the target. I wonder how working in a gravity field affects the process compared to orbital forces.
Perhaps you are saying that our ‘nauts simply don’t get the tens of thousands of practice those operators get. They also are not commensurately paid but that’s another subject.
Taking over some process when (apparently) needed might be useful, or save the bacon, in the short run; but those very same events are thereafter accounted for in the software, one imagines.
Dunno. I’m just a fanboy commenting on a blog, so don’t think for a picosecond that I don’t appreciate the training, effort, – and judgment – required of those brave people. Certainly they are in that dangerous and high-risk environment for a reason.
But on the other hand, has there been a situation in which some automated task went awry, only to be recovered by a human? None come to mind immediately but I will look around now to satisfy my curiosity. Story’s wrestling with Hubble’s balky doors comes to mind; there are other stories where the astronauts required a crafty switch of tools or procedures unpracticed in the pool.
And in this vein – sorry for the longish post – the stunning success ofJuno reminded us of just how good programming has become. Different problem, certainly, and just as surely a more difficult one.
Found this when I went searching. Neat site.
http://spaceflight.nasa.gov…
Not many ‘last-minute’ interventions, but quite a few incidents across the gamut that ultimately needed human intervention of some kind.
BTW, gravity definitely helps to make things somewhat easier when moving large objects around but it carries its own penalties. One gets used to dealing with zero-g dynamics (and orbit dynamic cross-coupling) with enough practice/training (even in just the simulators), but…it’s nice (in terms of necessary mental accounting) to be able to put something somewhere (say, down) while knowing that it won’t simply float off on its own independent trajectory when you let go of it.
But the fact that you can drop something (or fall & crash if you’re inside ‘it’) and thereby cause LOTS of damage quickly when gravity is present carries its own…weight…so to speak.
Good find…shows the importance of searching with just the right keywords.
Either “a judiciously-planned launch or an amazing amount of fuel” or lots of patience. The slower you take matching orbits and approaching, the easier it is. But there are practical limits to that. Most people’s patience is taxed by the ~1 hour it takes a commercial aircraft to go from the start of decent to arriving at the gate. Making the approach and docking phase for a spacecraft/space station a day-long procedure probably isn’t practical.
Perhaps not but as I’ve pointed out computers have a lot of patience.
Announcements via verbal communications invariably lag actual events. And, even during manual dockings selecting free drift can be part of an automatic sequence. The shuttle procedure had the crew confirm that their DAP (digital auto-pilot) transitioned to FREE after capture. They manually pushed a button to initiate the jet-firing sequence moments before contact, but that sequence finishes up by putting the shuttle into free drift.
But it sounds like you may have been closer to the mark anyway.