Our friend Hunchbacked is back

[ka9q]

I am going at him (again) on his ridiculous assertion that "whistleblower" engineers deliberately mounted the ascent stage propellant tanks asymmetrically to throw the LM off balance. His "proof" of this is the wallowing motion seen on the ascent movies and the statement in the manual that the RCS is needed to correct for off-axis thrust because the ascent engine is not gimbaled.

It was hard to know where to start with that one. Since he doesn't think Apollo is real, it's hard to see how he could think the mission films are evidence of anything. And he seems quite deaf to the fact that no matter how carefully one tried to align the center of gravity with the thrust axis, some imbalance would remain that would require active steering.

When I finally got him to understand that the oxidizer is 1.6 times heavier than the fuel, and that this would require asymmetric tank mounting to balance the stage, he retorted that the asymmetry was more than what was needed. So I found a good blueprint, brought it up in Gimp (similar to Photoshop) and carefully measured the tank positions. The fuel tank was exactly 1.6 times as far from the ascent engine centerline as the oxidizer tank. And now he says he needs to look for himself.

The guy seems to have no sense of real-world engineering at all. Or of the real world, for that matter.

[JayUtah]

And he seems quite deaf to the fact that no matter how carefully one tried to align the center of gravity with the thrust axis, some imbalance would remain that would require active steering.

This baffles me no end.  In real-world spacecraft engineering, it is a given that mass distribution alone cannot achieve practical passive stability.  It's a goal so unrealistic it's not even attempted.  Which is to say, we design spacecraft with passive stability in mind.  But first-order effects such as the depletion of consumables and second-order effects such as mechanical articulation simply cannot be made perfectly passive.  And the control techniques go right down to third-and-greater order effects such as elasticity and resonance in the spacecraft structure.  No spacecraft has been designed to achieve stability purely by passive means.  At best you need some kind of spin stabilization.  Usually you need moment-generating machinery, attitude sensors, and a closed-loop control system.  For practical manned space flight including rendezvous, these must be robust and capable systems.

Conversely, control design has never required "perfect" organization and distribution of control-moment generators.  There is no perfect placement for RCS jets such that you get no residuals.  As such, the mathematics for control system design have been fully generalized since the early 1960s and remain so today.  By "fully generalized" I mean based on linear algebra methods such that any combination of attitude errors and rates, and any combination of conjugate control inputs can be reckoned using the same generalized formulas regardless of actual direction or magnitude.  This gives rise to reliability engineering in the form of deliberately off-axis and/or non-orthogonal moment generators that tolerate the failures of single units (e.g., reaction wheels or jets).  Apollo had a limited ability to do this.

In other words, he's coming at the problem from someone who has some reasonable understanding of the basic dynamics problem, but who quite clearly has no experience whatsoever in the actual design and construction of spacecraft.

The guy seems to have no sense of real-world engineering at all.

Agreed.  This is why I'm skeptical of his education claims.  He may indeed have some sort of diploma, but it's abundantly clear he's never worked in the industry in any country.  And I think he wrongly believes people wouldn't be able to tell that.

[ka9q]

It's not like this stability/control stuff is intuitively obvious without some study or education. IIRC, even Robert Goddard fell for the usual misconceptions when he put his engine nozzles at the top of his rockets hoping that the rest of the mass would automatically hang below it like a pendulum. At least he eventually figured it out, though his obsession with secrecy probably made him take longer than it would have otherwise.

[bknight]

I am going at him (again) on his ridiculous assertion that "whistleblower" engineers deliberately mounted the ascent stage propellant tanks asymmetrically to throw the LM off balance. His "proof" of this is the wallowing motion seen on the ascent movies and the statement in the manual that the RCS is needed to correct for off-axis thrust because the ascent engine is not gimbaled.

It was hard to know where to start with that one. Since he doesn't think Apollo is real, it's hard to see how he could think the mission films are evidence of anything. And he seems quite deaf to the fact that no matter how carefully one tried to align the center of gravity with the thrust axis, some imbalance would remain that would require active steering.

When I finally got him to understand that the oxidizer is 1.6 times heavier than the fuel, and that this would require asymmetric tank mounting to balance the stage, he retorted that the asymmetry was more than what was needed. So I found a good blueprint, brought it up in Gimp (similar to Photoshop) and carefully measured the tank positions. The fuel tank was exactly 1.6 times as far from the ascent engine centerline as the oxidizer tank. And now he says he needs to look for himself.

The guy seems to have no sense of real-world engineering at all. Or of the real world, for that matter.

Another point in his video the RCS wasn't powerful enough to change orientation or abate those oscillations.

[ka9q]

Another point in his video the  wasn't powerful enough to change orientation or abate those oscillations.

The RCS was plenty powerful enough to overcome the residual imbalances and point the ascent engine as it needed to be pointed.

The "wallowing" motion so apparent in the ascent movies is caused by a "dead band" in the control laws. Instead of trying to hold attitude precisely in the desired direction, the computer fires the RCS only when the attitude drifts off nominal by some predetermined amount. Then it fires the RCS long enough to push attitude off-nominal by an equal amount in the opposite direction. The spacecraft swings back and the cycle repeats.

My understanding is that this is done to use RCS propellant more efficiently. They can't be throttled, but they can be fired in very short bursts to give you some desired average thrust over time. But it takes a finite time for the valves to open and for the propellants to flow and ignite, so even though you can command a very short burst it is quite inefficient because most of the propellants come out unburned. By firing longer and less often, more of the propellants are usefully burned. The small errors in attitude have negligible effect on the efficiency of the main ascent engine, and the autopilot can easily steer out any small course deviations that result.

[JayUtah]

The LM RCS was sized for the fully-docked, fully-fueled lunar module.  The notion that it wouldn't be able to control the relatively empty, relatively light ascent stage only is absurd in the extreme.  These are 100-lbf Marquardts on outriggers designed to lengthen the moment arm, rigged so that four jets could operate in concert for each roll, pitch, or yaw moment.  They were so oversized for this flight stage they often had to be operated in pulse mode.  Ed Mitchell described flying the ascent-only stage as "sporty."

[ka9q]

I wonder if thrusters would still be optimal for an LM designed with today's technology. An alternative would be a set of control-moment gyros for fine attitude control plus one or more thrusters used only to offload them periodically.

You'd still need thrusters for translation, e.g., during docking.

I guess it depends on how long the mission has to be, and particularly if the vehicle were designed to be reusable (e.g., refueled with propellants manufactured from lunar materials). The longer the mission, the more propellant could be saved by a set of gyros and the more attractive they look vs pure thruster control.

[JayUtah]

My understanding is that this is done to use RCS propellant more efficiently.

Yes, but in addition there are many reasons in control theory to allow a deadband, especially -- as you note -- when the system has a certain measurable latency.

The small errors in attitude have negligible effect on the efficiency of the main ascent engine, and the autopilot can easily steer out any small course deviations that result.

They average out.  If the spacecraft spends half its time on the negative side of the deadband and half its time on the positive side, the instantaneous attitude errors nominally integrate out to zero.  It's like changing lanes every 100 meters on the freeway (or dual carriageway).  You're constantly driving a few degrees either left or right of the centerline, but on average you're going down the road.  The initial ascent was allowed to have quite a wide dispersion, both in up-/downrange errors for LOI, and out-of-plane errors.  The idea for the initial ascent was to get to any orbit.  Then once a semi-stable orbit had been achieved, phasing maneuvers would bring the ships into compatible orbit.  In an extreme emergency, the CSM had some 17 contingency orbits planned to swoop down and rescue the (presumably disabled) LM.

[JayUtah]

Reaction wheels are preferred for ships that have a good supply of electrical power.

[bknight]

They average out.  If the spacecraft spends half its time on the negative side of the deadband and half its time on the positive side, the instantaneous attitude errors nominally integrate out to zero.  It's like changing lanes every 100 meters on the freeway (or dual carriageway).  You're constantly driving a few degrees either left or right of the centerline, but on average you're going down the road.  The initial ascent was allowed to have quite a wide dispersion, both in up-/downrange errors for LOI, and out-of-plane errors.  The idea for the initial ascent was to get to any orbit.  Then once a semi-stable orbit had been achieved, phasing maneuvers would bring the ships into compatible orbit.  In an extreme emergency, the CSM had some 17 contingency orbits planned to swoop down and rescue the (presumably disabled) LM.

You can clearly see that in the A17 Ascent and then he complains about the controller not centering the LM in frame, as it drifts to the left I think he assumes that the trajectory must be along the same direction as the camera is facing without rotating.  One other aspect, he seems to speed up the videos when is attempting to present his case. ie. A15 landing with the DAC camera.

[JayUtah]

I think he assumes that the trajectory must be along the same direction as the camera is facing without rotating.

It was expressly otherwise, and easy to find in the literature.  The LM ascended straight up for 10 seconds to avoid local terrain, then began a programmed sequence of increasing pitch-forward maneuvers.

[bknight]

I think he assumes that the trajectory must be along the same direction as the camera is facing without rotating.

It was expressly otherwise, and easy to find in the literature.  The LM ascended straight up for 10 seconds to avoid local terrain, then began a programmed sequence of increasing pitch-forward maneuvers.

My bad, in his information panel this is what he states "it is very clear that the plane of its ascent is perpendicular to the plane of the camera. The trajectory of the lunar module remains in this plane for all the initial the phase.  The LM has tilted nearer horizontal as we are looking at the engine.   it is right after this that the LM is yawing to the left from the perspective of ther camera I know you don't like going to YT, but if the description isn't understandable here is the linkhttps://www.youtube.com/watch?v=KAmQVZUUosg around 10:21

[JayUtah]

I may look at it later, but the fact also remains that the 16mm camera's line of sight was neither aligned along the line of flight nor oriented with respect to any LM axis.  This is most apparent in the descent videos (using the same camera mount) where one must tilt one's head to the side to see it as it might have appeared to someone standing at the LMP's station.

That said, there are also out-of-plane corrections done on the ascent.  At a certain fuel-optimal point, the LM could be programmed to "yaw" left or right (not strictly a yaw, but the informal equivalent) to change the plane of the orbit.  Basically Hunchbacked seems to have little understanding for what the ascent was meant to accomplish.  He's comparing the data to what sts60 accurately terms a "cartoon view of the world."

[bknight]

I may look at it later, but the fact also remains that the 16mm camera's line of sight was neither aligned along the line of flight nor oriented with respect to any LM axis.  This is most apparent in the descent videos (using the same camera mount) where one must tilt one's head to the side to see it as it might have appeared to someone standing at the LMP's station.

That said, there are also out-of-plane corrections done on the ascent.  At a certain fuel-optimal point, the LM could be programmed to "yaw" left or right (not strictly a yaw, but the informal equivalent) to change the plane of the orbit.  Basically Hunchbacked seems to have little understanding for what the ascent was meant to accomplish.  He's comparing the data to what sts60 accurately terms a "cartoon view of the world."

chuckle, it definitely yaws before the camera pans away, it may be programmed or corrective you know better than I

[ka9q]

It was expressly otherwise, and easy to find in the literature.  The LM ascended straight up for 10 seconds to avoid local terrain, then began a programmed sequence of increasing pitch-forward maneuvers.

It also performed a yaw maneuver during the vertical ascent phase to place the +Z axis in the desired orbital plane.

The main purpose of the vertical rise is to get some upward velocity to clear terrain before heading downrange. Pitchover is a rather abrupt 52 degrees.

If the LM doesn't start level, this isn't corrected until about 2 seconds after liftoff. This is most visible in the Apollo 15 ascent; you can clearly see the ascent stage initially moving to the left as it rises.