B-29 Superfortress Gun System Controls?

[IdahoRenegade]

I've always been interested in the gun system on the B-29. I understand that they were all remotely controlled from a central aiming station, or stations. Does anyone have any good sources of information on just how they worked, how the aiming systems were interfaced, and how effective they were, compared to say B-17 or B-24 manual systems. Or for that matter, how the remote turrets were controlled on those planes. I've done some searching, but what I've come across hasn't been very satisfactory.

Thanks.

 

[FLYBOYJ]

https://www.warhistoryonline.com/war-articles/b-29-remote-control-turret-system.html

 

[Joel Dobson]

When XXI Bomber Command on the Marianas (LeMay) ordered the guns removed from the B-29s on low level missions, what was removed? The barrels and the ammo? Or the entire turrets of the remote control system? Would it even be possible to remove the turrets??
If it was "just" the guns and ammo removed, how much weight was actually saved? Apparently, no one was really checking to see if a crew had them or not.
And how many rounds of .50 calibre was aboard a typical B-29 mission?

 

[FLYBOYJ]

When XXI Bomber Command on the Marianas (LeMay) ordered the guns removed from the B-29s on low level missions, what was removed? The barrels and the ammo? Or the entire turrets of the remote control system? Would it even be possible to remove the turrets??
If it was "just" the guns and ammo removed, how much weight was actually saved? Apparently, no one was really checking to see if a crew had them or not.
And how many rounds of .50 calibre was aboard a typical B-29 mission?

All the turrets (to include the guns) and I believe the actual fire control system. The tail gunner remained.

 

[Greyman]

Interesting video uploaded by Periscope Film not too long ago:

 

[Joe Broady]

I've always been interested in the gun system on the B-29. I understand that they were all remotely controlled from a central aiming station, or stations. Does anyone have any good sources of information on just how they worked, how the aiming systems were interfaced, and how effective they were, compared to say B-17 or B-24 manual systems. Or for that matter, how the remote turrets were controlled on those planes. I've done some searching, but what I've come across hasn't been very satisfactory.

Thanks.

The B-29 central station fire control system had five sighting stations: nose, upper, left blister, right blister, and tail. There were also five turrets: upper forward, lower forward, upper rear, lower rear, and tail. Each station could control one or more turrets in a priority system. A sighting station had primary control of a turret when the gunner at that station could obtain control of the turret at any time. A station had secondary control of a turret if the gunner could obtain control only if the primary control of the turret was released.

The nose station had primary control of both forward turrets. A control box had IN/OUT switches for these turrets to release control of one or both. This station had no secondary control of any turret.

The upper station had primary control of the upper rear turret, and secondary control of the upper forward turret. The blister stations shared primary control of the rear lower turret, depending on a LEFT / RIGHT switch in reach of both men. They had secondary control of the lower forward and tail turret. This could get complicated. One man could fire the lower forward turret in secondary control while his partner on the other side fired the lower rear and tail turret in mixed primary and secondary control. There were even switches for the blister gunners to select machine guns only, or guns and cannon when firing the tail turret.

The tail station had primary control of the tail turret and no secondary control of any turret.

Each sighting station had a reflector sight which displayed a reticle focused at apparent infinity. Before combat, the gunner would set the wingspan of the expected attacker on a dial. When the enemy appeared, the gunner put a central "pipper" (dot) on the center of the target and tracked it smoothly. At the same time he adjusted a range wheel to expand or contract a circle of dots around the pipper so they matched the wingspan. This was similar to the operation of a lead computing sight on a fighter, but the B-29 sights did no computing. They just transmitted sighting data to the computers, which were below the floor in the rear compartment (except the nose computer which was near the navigator).

Computer inputs and outputs were electrical, but internally the calculations were performed mechanically. The factors taken into account included lead due to target motion, drop due to gravity, and windage. That last correction depends on airspeed, altitude, and air temperature, so there was a panel for the navigator to set the values. And because the sights were remote from the turrets, a parallax correction was required. This is much like harmonizing a fighter's wing guns. They have to be angled in slightly to converge on the line of sight at the desired range. On the B-29 this correction had to be computed in real time due to the large distance from sight to turret and the different directions from which attackers could approach.

Amplidyne generators stepped up the computer outputs to power the two 0.5 hp drive motors in each turret. Each turret had provision for mounting a GSAP (gun sight aiming point) camera. An "overrun" control maintained camera operation up to 3 seconds after firing ceased in order to record the results of the burst.

Reference:
T.O. No. 11-70A-1, "The Central Station Fire Control System," 5 December 1944.

 

[Joe Broady]

Popular Science article from Feb 1945:
Secrets of B-29's Gun Controls

And this article from Jul 1945 explains a crucial component of a lead computing fire control system: the rate sensing gyro.
The Little Top That Aims a Gun

In that article is a photo of a B-29 gun sight. The entire upper part rotates left and right, and the central portion tilts up and down when you rotate the two notched handwheels. At the center of the left handwheel the round "action switch" is visible. When "on target" you close that switch with your palm. When slewing to a different position you release the action switch so the computer doesn't think you're tracking a target. Co-axial with the right handwheel (can't see it in the picture) is a slightly smaller wheel to control the diameter of the ranging reticle. On the frame, beside both wheels, are thumb buttons to fire the guns. On top is a gyro to sense angular rate, and a second gyro is beneath the optical portion that you look through. One gyro senses azimuth rate, the other, elevation rate.

Between the gyros, almost invisible in this photo, is the "retriflector" sight (mfd. by Bell-Howell). The gunner looks through an angled reflector plate and sees "a ring of illuminated dots forming a circle. An additional dot locates the center of the circle. The sight is equipped with two sky filters, one or both of which can be turned into the field of vision to emphasize the brilliance of the reticle image. The gunner, upon sighting a target, knows the size of some dimension of the target such as its wing span. If for example a target with a 60-foot wingspan is sighted the gunner turns the target size input knob until the figure 60 appears on the reflector plate. (The range of target size adjustment is 30 to 150 feet.) With this adjustment made, the gunner aims the sight at the target bringing the center dot of the reticle image to bear. Next he adjusts the diameter of the reticle image so that the circle of dots just spans the target. This is done by turning the range grip. This adjustment of the reticle image also serves to adjust a rheostat which supplies the computer with a signal proportional to the range. As the target comes closer the gunner keeps it constantly spanned." (Technical Order 11-70A-1, Handbook of Operation and Service Instructions, The Central Fire Control System, Model 2CFR55B1, 5 December 1944)

If the computer fails the gunner throws the computer switch to OUT. Then the turret(s) is simply slaved to follow the sight, and the gunner must estimate corrections.

All B-29 sights (except one) look like the magazine photo. In the nose the sight is mounted on a pantograph (an arrangement of pivoted linkages). "By this arrangement the vertical and horizontal axes of the sighting station always remain parallel to the corresponding axes of the airplane. Thus, without throwing sighting station and turrets too much out of their proper relationship, the gunner can move the sight to avoid obstruction of his vision by the ribbing in the nose structure. When not in use, the pantograph can be pushed against the right hand side of the ship and stowed securely."

The upper gunner station has a different sight because it's the only station able to rotate 360 degrees. This is accomplished by a rotating ring around the dome. The sight is attached to the ring with a mount which moves in elevation.

Regarding obstruction by the window frame, the gunner's manual says, "SIGHT WITH BOTH EYES OPEN KEEPING THEM SEVERAL INCHES BEHIND THE RETICLE. IT'S NATURAL, AND IT WORKS. Note to tail gunners: This is the nuts for shooting around tail window corner posts." (Gunner's Information File, General Electric GEJ-1631, not dated)

 

[Thumpalumpacus]

What a great post, J Joe Broady .

 

[Joe Broady]

And this article from Jul 1945 explains a crucial component of a lead computing fire control system: the rate sensing gyro.
The Little Top That Aims a Gun

That article shows the gyro held in a neutral position by a spring. When you continuously change the orientation of the sight, as when tracking a target, the precession force of the gyro overcomes the spring. The greater the angular rate, the greater the deflection from neutral.

In the B-29 it's different. Each gyro (there are two on each sight, driven at 10,000 rpm) is pivoted with a small degree of freedom and held in the neutral position by an erecting coil. The computer supplies power to the coil. As the gunner moves the sight (in azimuth, let's say), the erecting coil acts as an electromagnet to overpower the precession force of the gyro and hold it in the neutral position — if the voltage is properly proportioned to the azimuth rate. But suppose it's a little too low, so gyro drifts out of neutral. That closes an electrical contact in the gyro case, causing a motor to run inside the computer, which rotates a potentiometer to increase the voltage until gyro is pulled back to the neutral position and the contact opens. By this means the computer knows the rates at which the sight is rotating in azimuth and elevation.

However, the lead correction is not that simple. "Because the prediction correction is a function of both the angular velocity of the target and the time of flight required by the projectile, it is necessary to insert a resistance network in series with the erecting coil of each gyro. This is called the time of flight network and is not located in one unit. The range variable resistor is located in the range input unit [the portion of the computer which responds to the rangefinder wheel on the sight]. The density altitude resistors are located in the navigator's handset unit [where he inputs airspeed, altitude, and temperature]. The ballistic resistors are located in the ballistics calculating unit [within the computer]." (Technical Order 11-70A-16, Central Fire Control System Preliminary Instructions Operation, Service, and Overhaul for Computer Models 2HC1C1 and 2HC1D1, 1 August 1944)

Thus, the lead correction is calculated from the angular rates in azimuth and elevation, modified by the bullet time of flight. These corrections appear in the computer as shaft rotations of the azimuth and elevation differential selsyns. From the sight come selsyn signals representing the direction it's pointing. These go into the differential selsyns, where the computer's corrections are added. The sums are the desired azimuth and elevation of the turret in selsyn form, which are compared to the actual angles. Differences are amplified and cause rotation of the 0.5 hp turret drive motors.

I have omitted the parallax and ballistic (drop and windage) corrections. More on them later.