Rocks from the Moon

[JayUtah]

...the latter would have been the easier to develop.

Keep in mind that the Luna sample-return missions and the Apollo samples are apples and oranges. You have not addressed the discussion showing that retrieving 840 pounds of documented, differentiated, and specially treated samples is a markedly more difficult problem than retrieving 10 ounces of convenience samples.

[margamatix]

jayutah said:

[

How many automated systems have you personally designed and built? Please answer.

None, but I think you knew that.


Apologies to anyone I don't find time to answer everyone on every point by the way. Although I very much like this forum, I have two young children who are more deserving of my time, particularly as it is the school holiday here.

[JayUtah]

1. Rocket propellant. Very powerful for short durations. Useful only for generating thrust in rockets.

Well, to be honest you can harness rocket exhaust APU-fashion to generate non-propulsive mechanical energy. But again, only for a very short duration.

2. Batteries. Fairly powerful for short durations. Useful for powering drills, instruments, and rolling things like the LRV.

Funny you should bring that up. Remember how the core drills in Apollo got stuck all the time? The astronauts left them there to cool and went off to do other things and then came back to extract the tubes. If that were a core-sampling attachment on a rover, the rover would be stuck there, imprisoned in the lunar surface by its core sampler. That's why we don't put core samplers on unmanned probes. No problem for humans, but a serious problem for robots.

[sts60]

Neither scenario was possible at the time, but the latter would have been the easier to develop.

You have changed your argument. You said it was difficult but possible, and that the Soviets had done it a little later:
We have the technology to send unmanned vessels to the moon. Although it would have been extremely difficult to do in 1969, it falls within the realms of possibility that such a vessel could have collected rocks and returned to Earth with them at this time.

It's worth pointing out here though, that the USSR did not find it possible to do this until the Luna 20 project in 1972.
(Although they actually did it in 1970.)

The most difficult part of any moon landing, should we ever do one, will be in keeping the human beings alive.
That will always be the most difficult part of any manned landing on an extraterrestrial body, if you consider the mission as a whole.

But we have done one. Six, actually. We're still waiting for you to support your claim we haven't with something other than your opinions.

That's why Jay asked about your relevant experience - it's not that you have to be an aerospace engineer to contribute to the argument, but it does go to how seriously your opinion should be weighed when you say "this couldn't have happened" or "it must be this way", without ever saying why.

As far as not responding to every post, well, no one expects you to ignore your kids to answer seven or eight different people. But I do hope you listen to what we're saying. Despite all the political motivations behind the space race, the plaque on the Sea of Tranquility does say... "We came in peace for all mankind." Your children are heirs to that triumph, and it is meet that they be inspired by it.

[sts60]

I should have figured Jay wouldn't let the rocket-fuel thing slide by. Yes, I did think about it. I was just talking about what was actually used.

The way Jay can jump on stuff like that reminds me of the baseball line about trying to throw a fastball past a particularly good hitter who's "in the zone":

"It's like trying to throw a steak past a starving wolf."

Which, of course, is a compliment. ;-)

I didn't know about the drills getting stuck all the time. So I learned something new.

[JayUtah]

None, but I think you knew that.

Of course I knew that. I just want to keep things in perspective. I warned you that you can't handwave your way through this discussion; we are people who do automation, space travel, and human factors in engineering for a living. If you're going to start "talking shop" then you had better be prepared to give very detailed, very educated justifications for your opinions.

I could give you a whole list of the projects I've worked on, but I don't want to sound boastful. If you want to talk about my career and my background, I'm happy to oblige. If you just want to take at face value the claim that I am an experienced engineer, that suits me too. The point, either way, is that you can't bull---- us on these particular topics. This is what we do for a living.

You're trying to support your point of view by simply declaring true the premises on which it is based. In some cases those are simply indefensible declarations, made without any regard for the evidence or circumstances. In other cases the ones you declare true are the same ones you declare false in order to try to undermine someone else's point of view. These are the things that happen when you let some pre-desired conclusion guide your thinking.

Apologies to anyone I don't find time to answer everyone on every point by the way. Although I very much like this forum, I have two young children who are more deserving of my time...

That's fine, and enjoy your holiday.

I do try to get to each and every point that people bring up, but sometimes one slips by. The only consequence of your quite-appropriate priority on your family is that you can't really claim victory down the road if you've left all these points dangling. Your ability to put forward your opinion as objectively justified is exactly the extent to which you address points against it.

[Jason Thompson]

>>If it had all really happened, don't you think we would have landed and returned an unmanned craft before trying with a manned one?<<

If Concorde had really been mach 2 capable don't you think we would have sent an unmanned concorde up before putting people in it?

If the shuttle really worked don't you think we would have sent up an unmanned one before putting astronauts in it and sending them to space?

The people employed to fly these machines in Mercury, Gemini and Apollo (all of which were flown unmanned at least once before any human climbed inside them) were test pilots. It was their job to test new aircraft before they joined NASA, and it was their job to test the new space vessels when they were at NASA.

When all is said and done the best way to test the flight capabilities of a vehicle designed to be operated by a person is to put a person in it and get him to fly it.

Additionally, what reason would there be for believing the LM could not land? Apollo 5 showed that the machine worked in principle. Apollo 9 demonstrated that two men could fly it and that all its systems worked. Apollo 10 did the same in lunar orbit, and demonstrated that the crews could operate the LM safely on the far side of the Moon, out of contact with Earth, without risking a landing. Each time the LM was tested in incremental steps, first with no risk to any humans, then in Earth orbit and with the possibility of rescue by the CSM, then in lunar orbit with the possibility of rescue by the CSM. All that was left by Apollo 11 was to take the plunge and actually land.

By the time Apollo 11 came round it was known that the descent engine worked, the ascent engine worked, the RCS system worked, and the navigation radar systems worked. Surveyor had already demonstrated that soft-landing on the lunar surface was possible. What reason was there for not landing with Apollo 11?

A fully automated Apollo 11 would have landed in aboulder field and as a result very likely have crashed. Having a pilot on board solved the problem, and thus prevented NASA from wasting a perfectly good LM.

[JayUtah]

I should have figured Jay wouldn't let the rocket-fuel thing slide by. Yes, I did think about it. I was just talking about what was actually used.

Shuttle APUs immediately came to mind. They use pretty standard hypergolic rocket motors, not to propel anything but to generate electrical power and hydraulic pressure via turbines. Nothing would prevent such a system from being used on the moon to generate short bursts of, say, hydraulic power.

But the point I wanted to make was that of thinking outside the box. Often as engineers we have to step back and look at problems in the most abstract ways possible, shorn of meaningless constraints, to come up with new innovation. Doubtless the first cave man found that his rock hammer could also be a rock grinder. That's thinking outside the box. Nothing prevents me from hooking up a 100-lbf Marquardt motor to a turbine and pulling mechanical power off of it in order to run the pump on my lunar swimming pool.

The way Jay can jump on stuff like that...

Well, I get in that mindset when I'm at work. Part of my job is to make sure my branch of the engineering division doesn't do something stupid. it's not like my whole life revolves around finding errors in what other people do or say. But sometimes it's more instinctual than at other times.

I didn't know about the drills getting stuck all the time.

Black and Decker made the drills, using some very early NiCad or NiMH batteries. The bits kept getting stuck and B&D had to go through some design iterations in order to get the heat transfer on the motors working. The later missions had less trouble, but the first missions that tried to extract core samples had to do some pretty unorthodox things to get the core tubes in far enough to matter, and then back out.

[margamatix]

jasonthompson said:

>>
If Concorde had really been mach 2 capable don't you think we would have sent an unmanned concorde up before putting people in it?

If the shuttle really worked don't you think we would have sent up an unmanned one before putting astronauts in it and sending them to space?

.

I do see your point Jason and it's a very good one, but I'm sure you would agree that neither Concorde nor the Space Shuttle represented quite such a forward leap in technology as Apollo supposedly did.

[JayUtah]

I keep giving this same lecture. I should save it.

As I said, the goal was to put a man on the moon. If the goal had been to get moonrocks -- any moonrocks -- then an unmanned mission might have made more sense.

But if one of the objectives is to carry people, then you have to design for the people from the very beginning. And that means you can presume a smart pilot.

Why not test it first unmanned? Because it's not a valid test. If the design presumes a human pilot, then testing it without that pilot changes the conditions of the test. You have a "new" vehicle that will behave differently because it's automated instead of piloted, and you can't expect it to behave the same way as a piloted system.

Let's first be clear: the LM was tested unmanned. LM-1 flew a very simple mission in low Earth orbit under ground control augmented with some basic automation in its onboard computer (instead of the lunar landing programs). The test was generally successful -- so successful, in fact, that the follow-on test mission with LM-2 was cancelled, and LM-2 now sits in the Smithsonian. But a funny thing happened. When the DPS (descent engine) was fired up, the computer immediately shut it down. This happened repeatedly. The problem was that the ship had only been partially fueled in order to save weight at launch. With the fuel system partially depleted and operating at lesser pressure, the DPS developed less thrust than expected. The computer had been programmed to monitor the acceleration and to shut down the engine if the actual thrust was less than expected. This is important for pre-timed burns.

Here the test configuration was unfaithful to the flight configuration, and hence the test failed where the flight would have succeeded. The automation here did the wrong thing, although the motive behind it remains valid. A human pilot could have overridden the computer after learning what had gone wrong.

The one very important concept that non-engineers frequently fail to consider is that automation is not a universal savior. Automated spaceships are not "better" than piloted ones, and aren't guaranteed to succeed where piloted ships will fail. Instead, the reverse is usually true. Where the actual conditions vary from expectations (as they frequently do), the automation usually fails -- "goes off to La-La Land" -- whereas a human operator can apply intuition and save the day.

This is important because a crashed ship doesn't tell you much about why it crashed. An almost-crashed ship and its pilot are a wealth of information to help you refine the design. Eagle almost certainly would have crashed had Armstrong not taken over from the computer and flown the ship to a safe site. Because he did that we know how to solve his problem. If we had just lost contact with an unmanned Eagle when it augered in, we would have to guess at what killed it.

I'm sort of dancing around the important concepts, so I'll try to get to the point.

Take the bare bones of the LM: the structure and propulsion. Now in one case add a full automation suite, capable of flying a complex test mission unattended. In the other case add only the basic automation, such as was supplied in the Apollo LM. Because the two spacecraft contain fundamentally different control systems, they are two different spacecraft. Testing one will not allow you to understand the behavior of the other. It would be an unfaithful test.

Consider the two possibilities associated with testing through automation: the automation is generally more capable than the human, or the automation is generally less capable than the human. Then there is the cross-product with whether the test succeeds or fails, leading to the ordered pairs: (robust automation, success), (robust automation, failure), (inadequate automation, success) and (inadequate automation, failure). In some of the cases the tests would be considered valid. If supercapable automation fails, it is reasonable to conclude that a human pilot would also fail. Similarly if bad automation succeeds, we suspect a human pilot would also succeed.

It's what happens in the other cases that worries us. If robust automation succeeds, that's no guarantee that a human pilot will succeed. Your test will give a false confidence. If bad automation fails, you haven't shown that the problem cannot be solved, only that you cannot automate a solution. A human pilot may have succeeded where the automation fails. Since engineering is about solving problems by the best method, admitting failure in this case would be irresponsible.

The only valid test for a machine designed to be operated by a human is for a skilled human to operate it. No other test can be reliably generalized to the final product. Test validity is one of the concepts I'm dancing around.

The other is that of complexity effects. You might propose to solve the dissimilar control system dilemma by designing an LM that combines two control system methods -- one manual for operational flights and the other automatic for testing -- into one hybrid system. The two machines would then have identical control systems and you could possibly generalize the behavior of one to the other.

The unspoken (false) premise in this case is that unnecessarily built-in automation has zero "cost". I'm not talking about money, but rather the abstract notion of a detrimental effect on the design. Such a hybrid control system would necessarily be more complex than either aspect considered alone -- and more complex, in fact, than the mere sum of the complexities of each aspect. Complexity is bad because it leads to unexpected behavior. Thus systems are generally kept as simple as possible.

In this situation, a human pilot flying the LM by an overly complex control system might face control system states and behaviors that are detrimental to his safety and would not have been a problem in a simpler system. The LM with the hybrid system is inherently less reliable than either one of the simpler systems.

Further, the automation system was put in only for testing purposes. After the vehicle becomes operational there is no need for it. It serves no further useful purpose, and is there only to muddy up the works. One of the fundamental rules of engineering is that you don't introduce design features for obscure or limited purposes when those features have a detrimental effect on the common case.

As with many things in engineering, proper test methodology is not a matter of casual intuition. It must be carefully considered. Getting it wrong is not painless failure.

[margamatix]

jayutah said:

As I said, the goal was to put a man on the moon.

The goal was to divert the public's mind from VietNam.

[JayUtah]

...but I'm sure you would agree that neither Concorde nor the Space Shuttle represented quite such a forward leap in technology as Apollo supposedly did.

As a matter of fact I do disagree. The STS (shuttle) was a colossal leap forward over anything previously done. The Concorde perhaps not so much.

The Apollo command module was a direct descendent of the Mercury and Gemini spacecraft. The Saturn V was a scaled-up version of previous launch vehicles, hardened for manned use but containing no "grand leap" innovations.

The Apollo lunar module was a fairly "clean-sheet" design, but existed in a much cleaner problem space than the shuttle or the Concorde. The problematic constraints on the LM were size and weight. The actual operational objectives were not all that complicated.

[JayUtah]

The goal was to divert the public's mind from VietNam.

Hogwash. When Apollo was proposed in 1961, the war in Vietnam was going well. When the war turned bad in 1968 and subsequent years, the same people who were protesting the war were also protesting Apollo. They weren't distracted by it at all.

But you've unfortunately sidestepped the point. Apollo was intended from the start to be a manned program. It's whole intent -- or at least it's stated intent, if you wish to keep arguing distraction -- was to carry a man to the moon and bring him back again. Unmanned testing in that context has very limited usefulness.

[Bob B.]

margamatix said:

The goal was to divert the public's mind from VietNam.

When the goal was announced in 1961 there was no war in Vietnam; only a few advisors were in the country and no combat troops. How then could Vietnam be a motivating factor?

[Al Johnston]

margamatix said:

jayutah said:

As I said, the goal was to put a man on the moon.

The goal was to divert the public's mind from VietNam.

For the 1.5 seconds it took the newsreader to get to the next headline? ;D

Perhaps you should change your screen name to Tarquin Wim-bim-lim-tim F'tang F'tang Olé Biscuit Barrel