What degree of manual control did the Apollo astronauts have prior to TLI?

[VQ]

Flying the Saturn V during powered flight is a much different animal than on-orbit maneuvers after SECO.  Giving pitch and yaw commands from a hand controller via alternate command mode still washes those commands through the IU flight computer for necessary rate-limiting filtration, so you have very limited control anyway.  The problem with rocketry in general is that it's extremely easy to exceed the structural limits for bending moments.  You steer a rocket gently, and if gently isn't enough then all heads turn to the range safety officer.  The bending moment measured on the Saturn V (and I want to say it was at one of the joints at the S-IC/S-II interstage) was well over 100 kN·m in normal flight.  The last thing you want to do with a very large launch vehicle is fly it manually.

My recollection was that the level of possible pilot control was higher during the second and third stage burns than it was for the first, during which basically any critical failure would trigger automatic abort faster than the commander could react. Is there any truth to this recollection or am I just fabricating memories?

[ka9q]

That's almost certainly true, as second and third stage flight occur outside the atmosphere so angle of attack isn't a problem.

As you accelerate through the atmosphere the aerodynamic drag increases with velocity squared, but it also varies linearly with air density, which decreases exponentially with altitude. So it increases to a peak called Max-Q and then drops again. This occurred at 78.9 seconds and 37.5 kilopascals on AS-503 (Apollo eight). Sea level atmospheric pressure is about 101 kPa, so you can see this is quite a bit of pressure.

But it falls just as fast as it rises, and by staging at 154 sec at an altitude of about 60 km, air density and therefore drag has decreased to essentially zero.

Maximum bending moment on that flight occurred at 74.7 sec, just before Max-Q, and was 6.78 meganewton-meters. That sounds like a lot, but the design load is 30 MN-m. This occurs near the bottom of the LOX tank in the first stage (the LOX tank is above the RP-1 tank). Bending moments tend to vary a lot from flight to flight due to differences in high altitude winds. It's why they release so many weather balloons before a launch.

[smartcooky]

As you accelerate through the atmosphere the aerodynamic drag increases with velocity squared, but it also varies linearly with air density, which decreases exponentially with altitude. So it increases to a peak called Max-Q and then drops again.

Ok, so let me see if understand this correctly....

1. As the rocket accelerates from lift off, the velocity component of aerodynamic drag increases, and as its altitude increases, the air density component decreases.

2. In the first part of the flight, the velocity component increases faster than the air density component decreases, causing a nett increase the aerodynamic drag.

3. Eventually, the rocket reaches an altitude where the atmosphere is thin enough that the air density component begins to fall faster than the velocity component increases and that point is Max Q.

Wasn't it just after Max Q that Challenger exploded?

[ka9q]

The Emergency Detection System (EDS) was specifically designed to cause an automatic abort during the first two minutes of first stage flight when even small attitude errors can produce bending moments beyond what the structure can handle.

It used angle rates from the gyros in the IU and angle-of-attack sensors in the Q-ball (the very tip of the launch escape tower). It also watched for loss of electrical continuity between the CSM and IU, and for failure indications from the engines. A pitch or yaw rate exceeding 4 degrees/sec, or a roll rate exceeding 20 deg/sec would trigger an abort, as would the loss of 2 or more F-1 engines.

As far as I can tell, excessive angle-of-attack did not trigger an automatic abort, but it could be used as an indication for a manual abort. It was apparently measured by differential pressures sensed by the Q-ball. The abort limit was 22 kPa; the maximum reached on AS-503 was 4.8 kPa, again around Max-Q.

The crew disabled the EDS at about 2 minutes. They could still command a manual abort if they saw excessive rates. The pitch and yaw limits were relaxed to 9.2 deg/sec (from 4 deg/sec); the roll limit remained at 20 deg/sec. And of course the Q-ball was jettisoned along with the escape tower shortly into second stage flight.

[ka9q]

As you accelerate through the atmosphere the aerodynamic drag increases with velocity squared, but it also varies linearly with air density, which decreases exponentially with altitude. So it increases to a peak called Max-Q and then drops again.

Ok, so let me see if understand this correctly....

Yes, you have it right. Aerodynamic drag is 1/2 C_d rho A v². C_d is the coefficient of drag, A is the cross sectional area, rho is the air density and v is velocity. Air density decreases exponentially with altitude.

The path through the atmosphere is pretty much determined by keeping the angle of attack as close to zero as possible; this is done by flying a "gravity turn" where the rocket is momentarily pitched over shortly after clearing the tower, then the gimbals are restored to near 0 degrees. Then gravity does the rest. The guidance system uses predetermined values during this time. Around staging, attitude is held constant. After staging, when the rocket is in vacuum, the guidance system is allowed to operate "closed loop" by steering it from where it is (in position and velocity) to where it wants to be.

Wasn't it just after Max Q that Challenger exploded?

Yes, but that was probably coincidental. It happened when the plume through the breach in the SRB had eaten away enough of the rear SRB strut that it was free to swing away and allow the nose to pivot into the LOX tank at the top. The bottom of the LH2 tank also fell away.

[Northern Lurker]

Wasn't it just after Max Q that Challenger exploded?

Yes, but that was probably coincidental. It happened when the plume through the breach in the SRB had eaten away enough of the rear SRB strut that it was free to swing away and allow the nose to pivot into the LOX tank at the top. The bottom of the LH2 tank also fell away.

AFAIK on SRB ignition sections of primary and secondary O-rings were burned away but fortuitously an oxide plug formed and sealed the leak. After Max Q the stack encountered wind shear which caused bending moment which dislodged the plug and allowed hot gases to impinge the support strut and the External Tank.

Lurky

[gwiz]

Ok, so let me see if understand this correctly....

There's an extra complication in that the drag coefficient isn't a constant.  It increases with angle of attack and is also a function of Mach number, most marked by a sharp increase as the Mach number approaches and passes through 1.

[bknight]

AFAIK on SRB ignition sections of primary and secondary O-rings were burned away but fortuitously an oxide plug formed and sealed the leak. After Max Q the stack encountered wind shear which caused bending moment which dislodged the plug and allowed hot gases to impinge the support strut and the External Tank.

Lurky

From what I remember a piece/chuck of solid rocket propellant was the temporary plugging agent

[Northern Lurker]

From what I remember a piece/chuck of solid rocket propellant was the temporary plugging agent

On the morning of the disaster, the primary O-ring had become so hard due to the cold that it could not seal in time. The secondary O-ring was not in its seated position due to the metal bending. There was now no barrier to the gases, and both O-rings were vaporized across 70 degrees of arc. Aluminium oxides from the burned solid propellant sealed the damaged joint, temporarily replacing the O-ring seal before flame passed through the joint.

[bknight]

I stand corrected.

[ka9q]

Yes, that's correct; propellant combustion residue temporarily blocked the breach in the O-rings. As I recall, it was upper altitude wind shear that eventually broke the residue and let the plume re-emerge. It also took some time for that plume to burn through the strut holding the SRB to the ET and to cause the bottom of the LH₂ tank to fail.

Speaking of max-Q, you can really hear the effect in the videos made within shuttle cabins during launch. At liftoff there is a lot of noise and shaking, as you'd expect. That settles down, but as the shuttle accelerates the wind noise grows very loud, reaches maximum at max-Q, and then tapers off again. It becomes eerily quiet in the cabin even before the SRBs burn out and are jettisoned, which is accompanied by a noticeable bang and a flash in the windows from the separation rockets.

[JayUtah]

Maximum bending moment on that flight occurred at 74.7 sec, just before Max-Q, and was 6.78 meganewton-meters.

Thanks for the correction; you obviously looked up the values while I was trying to remember it.  The value I had in my head was for a different parameter of the Saturn v.

Bending moments tend to vary a lot from flight to flight due to differences in high altitude winds. It's why they release so many weather balloons before a launch.

And indeed, as you allude to below, why post-Challenger launch criteria were amended to add more winds-aloft data.

As I recall, it was upper altitude wind shear that eventually broke the residue and let the plume re-emerge.

You recall correctly.  The steering moments generated in response to the wind shear flexed the SRB more than usual and cracked the oxide plug loose.  Regardless of how it may have behaved on the Challenger flight, engineers realized that their design limits for flexure on the field joints were not nearly conservative enough given the emerging knowledge of the joint's dynamic behavior, and this led to more stringent wind-shear requirements for launch.  This had the effect of limiting the anticipated steering moments to those within the existing experience base, excluding Challenger.

Speaking of max-Q, you can really hear the effect in the videos made within shuttle cabins during launch. At liftoff there is a lot of noise and shaking, as you'd expect. That settles down, but as the shuttle accelerates the wind noise grows very loud, reaches maximum at max-Q, and then tapers off again.

This is in part a response to discontinuities in the forward section of the orbiter.  The angle between the nosecone and the windscreen "traps" air, and the inset of the windscreen panes certainly doesn't help any with that behavior.  There's a lot of dynamic pressure applied to the windscreen, and the inset produces turbulence.  That's a huge roar just a meter or so from your face if you're in the commander or pilot seats.  Boeing struggled with a similar problem for years; their airliner flight decks were notoriously noisy, for many of the same reasons.  Even as late as on the 777, the particular way in which they traditionally formed the forward section of the fuselage included these discontinuities -- the result of structural methods to frame the windscreens.  On the 777 tabs were added to the top surface of the nosecone in an attempt to redirect the airflow to the sides of the windscreen.  Boeing wasn't responsible for the orbiter cockpit design, but the orbiter designers wrestled with similar problems and design constraints.  With the 787 Dreamliner, the entire structural system of the forward section was redesigned from scratch and we were able to entirely eliminate the "Boeing roar" on the flight deck.  The 787 windscreen is entirely conformal to the overall fuselage shape.

Also on the orbiter, the flow did not "stick" very well to the top of the flight deck as it came off the windscreen.  There was intermittent flow separation right above/behind the windscreen that would have produced a fair amount of turbulence there as well, and that's yet another source of noise.  The extremity of this behavior can be seen in this photo of the stack as it transits the sound barrier.

[Count Zero]

The extremity of this behavior can be seen in this photo of the stack as it transits the sound barrier.

...and I have a new desktop background!

The Emergency Detection System (EDS) was specifically designed to cause an automatic abort during the first two minutes of first stage flight when even small attitude errors can produce bending moments beyond what the structure can handle.

It used angle rates from the gyros in the IU and angle-of-attack sensors in the Q-ball (the very tip of the launch escape tower). It also watched for loss of electrical continuity between the CSM and IU, and for failure indications from the engines. A pitch or yaw rate exceeding 4 degrees/sec, or a roll rate exceeding 20 deg/sec would trigger an abort, as would the loss of 2 or more F-1 engines.

I read a report a couple of years ago that concluded that if one of the outboard F-1s went-out during ascent, vehicle destruction due to bending moments would occur faster than the LES could fire.  This makes some sense to my layman's mind:  If I imagine the Saturn V stack as a 30-story building stressed with 4 gravities, and then one of the corner supports fails, I would think Really Bad Things will Happen Really Fast.

Using more formal terms, the report went on to predict that the spacecraft would be flicked from the tip of the rocket like whipped-cream from a spoon.  In looking for ways to mitigate the lateral forces on the spacecraft, the report ironically pointed-out that these forces would be substantially less if the escape tower wasn't there providing leverage.

[ka9q]

Was that before or after the practice of "canting" the engines was begun? With canting, the four outboard engines are gimbaled slightly outward so the thrust vector from each one runs through (or at least closer to) the c.g. of the stack. The purpose is to minimize the disturbance if one of those engines should suddenly fail.

[Peter B]

...The angle between the nosecone and the windscreen "traps" air, and the inset of the windscreen panes certainly doesn't help any with that behavior.  There's a lot of dynamic pressure applied to the windscreen, and the inset produces turbulence.  That's a huge roar just a meter or so from your face if you're in the commander or pilot seats.  Boeing struggled with a similar problem for years; their airliner flight decks were notoriously noisy, for many of the same reasons.  Even as late as on the 777, the particular way in which they traditionally formed the forward section of the fuselage included these discontinuities -- the result of structural methods to frame the windscreens.  On the 777 tabs were added to the top surface of the nosecone in an attempt to redirect the airflow to the sides of the windscreen.  Boeing wasn't responsible for the orbiter cockpit design, but the orbiter designers wrestled with similar problems and design constraints.  With the 787 Dreamliner, the entire structural system of the forward section was redesigned from scratch and we were able to entirely eliminate the "Boeing roar" on the flight deck.  The 787 windscreen is entirely conformal to the overall fuselage shape.

Also on the orbiter, the flow did not "stick" very well to the top of the flight deck as it came off the windscreen.  There was intermittent flow separation right above/behind the windscreen that would have produced a fair amount of turbulence there as well, and that's yet another source of noise.  The extremity of this behavior can be seen in this photo of the stack as it transits the sound barrier.

Jay

I seem to remember reading somewhere that the flow of air between the Orbiter and the ET turned out to be more dangerous than designers had expected, and this alone might have caused the destruction of Columbia in the first Shuttle mission.

Is this true or am I mistaken?

With only a lay knowledge of aerodynamics I get the impression that some of the air passing off the nose of the ET would funnel between the Orbiter and ET, and in the process get compressed as the space narrowed, and on top of that be disrupted by the struts and pipes connecting the two. Would it be right to say that around the time of Max Q that airflow would be doing a pretty good job of trying to shake the two apart?