Mechanisches Rechnen mit präzis geformten Stahlkörpern und Zahnrad-Getrieben

Complete English translation of the original German-language document (22 pages).

Mechanical Computation with Precisely Formed Steel Bodies and Gear Trains

The Museum of Communication holds 143 negative foils (approximately A4 size) containing the results of calculations performed by the firm HASLER, Bern, which were evidently delivered to Italy. They are numerical instructions for milling depths used in manufacturing the ballistic form bodies (cams) in the fire-control device for the Italian anti-aircraft guns 90/53. All foils have been examined; individual ones were photographed freehand from a light table. The principle of computing with form bodies is presented here. An outlook sketches how the data for computing the form bodies were presumably obtained, with questions relating to air resistance.

Three appendices:

Appendix A. The data on the HASLER foils do not match the known HASLER command devices — p. 16
Appendix B. Air resistance of projectiles and the “Mariandl” integraph, with a description of the ball integrators — p. 17
Appendix C. Chronological development and innovations in the HASLER command devices — p. 20

Dating / Names: The foils are only very sparsely dated. Calculations: 1950. Individual corrections: 1951. The sheets are only rarely signed. One M. BERCHTEN signed them, not the expected E. SCHULTZE. Individual short signatures on corrections are indecipherable.

Relation to the contemporary HASLER devices: The anniversary publication HASLER — 1852–1952 (but not on the internet) contains the statement that HASLER could supply its own command devices to Italy. The form bodies corresponding to the documented calculations do not, in part, match the known HASLER devices used in the Swiss Army. It appears that HASLER undertook further development of its existing devices for the Italian export (see Appendix A). From the years 1950 and 1951 onwards, experiments and adaptations were made for the Swiss devices in order to process radar data (two angles, distance) with the existing mechanical command devices. The HASLER foils fall precisely within the timeframe of these adaptations. The Italian command devices B.G.S. had been using radar data experimentally from as early as 1942/43 (Würzburg, Volpe), and after the war with the British Mark II. Later, Contraves fire-control devices F90 BT were employed. Contraves Italia manufactured approximately 300 F90 fire-control devices for NATO countries (electronic analog computer with tubes and rotating shafts, single-axle trailer “Fledermaus,” precursor to the “Super-Fledermaus”).

Description of the computer GAMMA-JUHASZ, later GAMMA-HASLER: Second work at http://e-collection.library.ethz.ch/eserv/eth:47590/eth-47590-01.pdf

Edited: December 2015 to March 2016 by A. Masson, CH-4900 Langenthal (born 1946)

Distribution: Museum of Communication Bern, Aviation and Anti-Aircraft Museum Dübendorf, Library at Guisanplatz Bern, analogmuseum.org (Prof. Ulmann), History of Technology ETH Zurich, History of Technology Jena, Deutsches Museum Munich, Wehrtechnische Sammlung Koblenz, Deutsches Technikmuseum Berlin, ENTER Museum Solothurn, ISER Erlangen, Herbert Bruderer formerly ETH, and others.

Keywords: Mechanical analog computers / Heavy anti-aircraft artillery / Command devices / Ballistic bodies (English: ballistic cams; French: camoïde) / Form bodies / Air resistance / Ball integrator

Mechanical Computation with Form Bodies and Gear Trains

With precisely formed bodies, functions of two or three variables can be stored in fixed form. During computational operation, the bodies are continuously repositioned, and the currently valid function values are read out by tracing the surface. Numerical tables from the firm HASLER, produced for Italy in the years 1950/51, provide a glimpse into the world of mechanical analog computers of that era. All calculations for the form bodies originally had to be carried out by hand (for automatic trajectory calculation, cf. Appendix B, p. 17, Ballistic Integraph “Mariandl,” 1948). HASLER’s mathematical staff member, Dr. E. Schultze, was only able to compute on the first digital computer ZUSE Z4 at ETH for approximately 25 hours in 1955.

Task of the Fire-Control Computers

The heavy anti-aircraft guns (approximately 1930–1960) could only be employed thanks to a mechanical computing aid that determined the point of intersection of aircraft and projectile. The point of intersection was calculated from the position and velocity of the aircraft, as well as the flight time of the projectiles — which, however, could only be determined once the point of intersection was known. Everything is complicated by the curved trajectories of the projectiles and the strong reduction in their velocity due to air resistance. The computer continuously transmits to the guns the two angles to be set (azimuth and elevation) as well as the continuously changing values for the time fuze. All these values are set manually at the guns by means of handwheels and “follow-up pointers” according to the computer’s instructions. Typical projectile flight times are 10 to 20 seconds; computation is performed for even greater distances with projectile flight times of over 30 seconds.

Computational Principle

Two men track the aircraft through a telescope and supply the computer with the azimuth and elevation angle of the measuring point. A third sets the range values from the rangefinder (tubes 2 m to 4 m long, in Switzerland on their own tripod). From these, the mechanical computer determines all parameters of the flight movement such as altitude, horizontal distance, course angle, horizontal and vertical velocity, and — in the HASLER devices from 1943 onwards — also elements of curved flight. From the flight altitude and the horizontal distance, for given firing trajectories, it is in principle known how much higher the gun must aim (due to the curved projectile trajectory) than the elevation angle of the aircraft. These values have been stored for innumerable combinations of altitude and distance in a carefully milled “ballistic body.”

These are precisely ground steel bodies that can be shifted along their central axis by a cable drive on a carriage (corresponding to the map distance), and additionally rotated into a specific angular position by a gear (corresponding to the flight altitude). A tracer pin at a fixed location presses against each body and measures the local distance of the surface from the axis, yielding, for example, the current required projectile flight time. The bodies are continuously shifted and rotated as the aircraft’s path changes (a real-time mechanical computer).

Inventory of Form Bodies in the Command Device GAMMA-JUHASZ (1938) and later GAMMA-HASLER (1940, 1943, 1950)

There are three large, precise form bodies:

a) From the horizontal distance to the point of intersection* and from the flight altitude, the flight time of the aircraft from the current surveying point to the detonation point is determined.

b) From the horizontal distance to the point of intersection* and from the flight altitude, the gun’s elevation is determined.

* if the shot were fired “now”

c) From the horizontal distance to the point of intersection (increased by the distance traveled by the aircraft during the gun’s loading time) and from the flight altitude, the flight duration of the shell is determined.

The distances are measured in the computer from a reproduced flight path to scale, taking into account the direction of flight and the calculated lead distance.

There are numerous additional, smaller form bodies: In the vertical as well as in the horizontal direction, the velocity of the aircraft is multiplied by the flight time to determine the lead. Both multiplications are performed with a form body.

The angles transmitted to the guns can still be subjected to daily corrections. This also applies to the time fuze setting, i.e. to the clockwork of the fuze. If systematic deviations are recognized, the values can be corrected and improved by hand. In this way, the initial velocity of the projectiles or the air density can also be entered with current values, as can longitudinal and lateral wind. For these fine corrections, a total of 14 smaller form bodies are provided. Each individual output value is a function of three quantities and therefore requires two form bodies. In total (apart from the direct manual corrections at the very end), there are three computed corrections to the time fuze setting, three computed corrections to the elevation angle, and one computed correction to the azimuth angle.

From the further developed device GAMMA-HASLER 1943 onwards, an additional multiplication is added, since HASLER (possibly for the first time in the world?) incorporated a curved-flight extrapolation. Multiplications could also have been achieved using lever gearing based on the intercept theorem (see page 11), but here form bodies are used exclusively for multiplications.

Description of the Large Form Bodies (Illustrations from the manuals)

Scale for the large “ballistic bodies”:

Longitudinal displacement: 1,000 m map distance corresponds to an axial displacement of the body of 12.5 mm (shifted by cable drive).
Rotation: 1,000 m flight altitude corresponds to a rotation of the entire body of 30° (rotation via gears).
Distance of the surface from the axis, flight time: 1 second corresponds to 1 mm (tracer pin, see figure at right).
Total length of the active surface approximately 13.6 cm.

[Image 303]

Reading out the flight time, for example: A tracer pin presses against the surface at a fixed location. The body moves away under the pin and rotates. The tracer pin operates an electrical center contact that controls the current to two electric motors, one lifting the frame upward, one downward (same shaft). The electric motors continuously follow the traced surface — the “output” of the shaft “flight time” for further computation is not drawn here (one of the four gears). The motors often also have brakes.

[Image 285]

Smaller Form Bodies in the “Daily Corrections”

After the computation is reasonably established (aircraft surveyed, point of intersection found, flight times determined, gun parallax accounted for), the values are further corrected according to wind, air density, powder temperature, as well as through three freehand, direct manual corrections for azimuth and elevation angle and fuze time, corresponding to the observation of fire.

[Page 3]

GAMMA-HASLER Model 1940

Corrections of the computed values immediately before output to the three electrical follow-up pointers for the guns (lower right, for azimuth, elevation, time fuze).

Seven smaller rotational bodies (labeled with letters) correct for air density (Delta), longitudinal wind (WL), initial velocity (Delta v0), and lateral wind (WQ). The corrections are applied as a function of projectile flight duration (shaft A) and flight altitude (shaft C). The rotational bodies are computed for the maximum value of the disturbance variable. Pressed against them sit the seven “reduction bodies,” which convert to the actual disturbance variable set by hand. The wind resolver (at the very top) splits the entered wind (two handwheels at top) into longitudinal and lateral components.

The corrections of the three stages are added by means of lever linkages and lead to the correction of the projectile flight time (upward) or the barrel elevation (downward), respectively. The lateral wind influences, via shaft D, only the azimuth angle alpha of the guns (correction in differential 85 and 5, added to the parallax correction from below). Shaft E brings the azimuth angle of the guns to the wind resolver (information “longitudinal” / “lateral”).

[Image 392]

Each of these 7 daily corrections depends on three variables. Additionally, the 3 dark knobs on the right allow the outputs to the guns to be adjusted by hand.

Two form bodies in cascade! Left body: correction of the fuze time or elevation at maximum possible v0, depending on flight altitude and projectile flight time. Right body: reduction to the set v0.

[Image 307]

Daily corrections wind, air density, v0. Mod. 1938, approx. 30 × 20 cm. 7 form bodies, 7 reduction bodies. At the bottom, the levers for adding three disturbance variables; above, the same with electrical micrometer.

[Image 264]

Description of the Historical HASLER Foils

In the Museum of Communication there are 143 foils with numerical results that HASLER delivered to Italy. They are negative foils of large tables, packed in four envelopes, containing the results of numerical calculations for five subjects. Envelope inscriptions and subjects:

Subject	Foils
Correzioni 90/53 — Scostamento — Spoletta pirica (2 sub-topics)	14 foils
Spoletta pirica	11 foils
Derivazione	12 foils
Elevation 90/53 — Inclinazione totale 90/53	46 foils
Durata 90/53	60 foils

Subject labels: Daily corrections — Time fuze — Derivation — Elevation — Aircraft flight time

The results of the calculations are only rarely dated (1950); corrections of individual values also from 1951. It is evident in multiple places that newer calculations have been entered by another hand for increased distances or increased projectile flight times (extension of the measurement range).

The notation 90/53 refers to the Italian anti-aircraft gun 90/53 (also anti-tank), with caliber 90 mm and barrel length 53 times the caliber. This gun entered service in 1939 or 1940 and was not retired until 1964 or 1970. It is reported to have fired for the last time in Montenegro in 1990. V0 approximately 850 m/s.

No notation whatsoever is found for the type of command device for which everything was calculated! Several of the tables are written on official, pre-printed HASLER paper. No information can be found on the internet to the effect that HASLER ever delivered its own command devices to Italy, as mentioned in the anniversary book “Hasler 1852–1952.”

HASLER promotional brochures were certainly sent to Italy, as to many other countries (Egypt, India), but whether a sale was concluded remains open. The 2005 brochure of the HASLER Foundation states that “various countries” became customers for the 1950 device. The 2005 brochure is evidently strongly (exclusively?) influenced by the 1952 anniversary book.

Italy used the following command devices:

Device	Period
GALA	1937–1943
GAMMA-JUHASZ	from 1940
BORLETTI-GALILEI-SAN GIORGIO B.G.S.	from 1940 development; from 1941 in service with gun 90/53, until approximately 1964

The B.G.S. command device was later also supplied with radar data: Würzburg from 1942, British Mark II in the 1960s. The 90/53 gun was also controlled using the CONTRAVES fire-control device F90 BT (see below).

Numerical Calculations Found in the HASLER Foils for These Form Bodies

Subject	Number of bodies	Variables (independent)	Total range	Step size	Computed values (all foils)	Values per foil (if filled)
Correzioni Zeit	7 small bodies	Time: 0–25 sec; Flight altitude: 0–12,000 m	0.2 sec / 200 m	—	3,600	—
Spoletta pirica	1 large body	Time: 0–30 sec; Flight altitude: 0–11,500 m	0.25 sec / 50 m	—	2,880	—
Derivazione	1 large body	Time: 0–30 sec; Flight altitude: 0–11,500 m	0.25 sec / 50 m	—	650	—
Inclinazione	1 large body	Flight altitude: 0–11,200 m; Map distance: 0–12,000 m	100 m / 50 m	—	464	—
Durata	1 large body	Flight altitude: 0–11,200/11,600 m (extended range 100 m); Map distance: 0–12,000 m	100 m / 50 m	—	480	—

Number of computed values: After some deduction for partially filled foils (where both variables approach their upper range limits and computation was discontinued), some 100,000 to 120,000 values may have been computed by hand or interpolated from tables.

Intended accuracy: One should picture a large bomber at many kilometers’ distance, flying at 6,000 m altitude — its wingspan is 30 meters, and the computation grid is nearly that fine (50 m each in altitude and distance)!

Computational effort: For every additional gun / for every new initial velocity, the entire work must be repeated!! In Switzerland, in addition to the war ammunition (v0 = 805 m/s), there was also special firing-range ammunition (v0 = 550 m/s), so that the lead for the much slower towed aircraft would be similar to that for war ammunition. The complete sets of ballistic bodies had to be calculated twice, manufactured twice — when needed, the command device had to be opened and fitted with another set of form bodies.

Comparison of the HASLER Tables with the Ballistic Bodies of the Known GAMMA-JUHASZ and GAMMA-HASLER Devices

1. The daily corrections appear to correspond well with the seven correction bodies according to the known gear diagram. An unusual and puzzling notation is used on the foils to designate the individual corrections. However, since notes are found showing how the HASLER staff “translated” these for themselves and scribbled in a more readable, conventional notation in small script, there can be hardly any doubt. — Only those bodies have been calculated (or are still present in the documents) that bear a letter in the diagram (cf. Image 392, page 4). Nothing of the reduction bodies pressed against them appears in the foils. They will all have the same form, i.e. their form converts a 100% disturbance linearly to the actual value.

2. Spoletta pirica (time fuze): This body definitively does not correspond to the known HASLER command devices, because the variables do not match. In the HASLER devices for the Swiss Army, the time fuze setting of the projectiles (time from firing to detonation) is determined from the flight altitude and the horizontal distance. A time variable as an input does not exist there.

The various time intervals used are somewhat delicate — the flight time of the aircraft from its current position to the detonation point is something other than the flight time of the shell. — What is unclear and rather strange about these foils is the fact that negative values are also recorded as results.

3. Derivazione (projectile deflection due to rifling and gyroscopic effect): In the HASLER devices, there is no known record of there ever having been a dedicated derivation form body. Even in the most recent device 43/50 R with the radar intermediate deck (prototype 1950, in service from approximately 1953), nothing of the sort is noted in either the gear diagram or the electrical diagram. The derivation appears to have been accounted for by forming the long tracer needle on the measuring carriage (which indicates the flight angle and the map distance) not straight, but slightly curved. The lateral deviation is thereby coupled to the horizontal distance (rather than to time), which represents an approximation. M. BERCHTEN signed precisely those number sheets that do not match the HASLER devices: Spoletta pirica and Derivazione. See also Appendix A, page 16.

4. Inclinazione, and 5. Durata (flight time measured at the aircraft): In the HASLER devices these two form bodies sit on the same carriage and are driven by the same horizontal distance and the same flight altitude. The foils appear to match, in terms of their variables, the ballistic bodies of the previous HASLER devices. However, the scale does not match exactly: an additional 1,000 m of flight altitude corresponds to 30° of rotation angle for the body in GAMMA-HASLER, but only 28° (at 7,000 to 9,000 m) in the foils for Italy, and in one foil at very low range (below 1,000 m) even only 27.5°. Different again for Durata (flight time): the foils specify 30° of rotation per 1,000 m of flight altitude.

[Page 6]

In summary, it must be said that two of these form bodies do not correspond to the known HASLER devices. Whether there was a further development for Italy (which may no longer have been produced) remains unknown. The Museum of Communication has a modified HASLER device 43/50 R that, at the output synchros, looks clearly different from the devices of the Swiss Army (cf. p. 17). In purely chronological terms, the foils (1950/51) fit well with the development of the HASLER device 43/50 R with the radar intermediate deck.

The Italian B.G.S. device is less likely a candidate, since the table of contents of the operating manual suggests a stronger emphasis on Cartesian coordinates, which are not used at all in the HASLER devices.

Perhaps the form bodies were calculated specifically for the Italians, but then the HASLER devices were ultimately not delivered? The claim that a successor to the GAMMA device “through CONTRAVES” was delivered to Italy may be incorrect:

“Successivamente verrà utilizzata una nuova centrale di tiro derivata dalla GAMMA, la F/90° B della CONTRAVES, asservita ai radar tiro e di scoperta, forniti da UK e CANADA, l’AA N3 Mk. 7 e l’AA N4 Mk. 6/2, rimanendo in servizio fino ai primi anni ‘60, affiancando i materiali da 90/50 e 94/50 forniti dagli USA.” (See Sources, page 15, second address from Italy)

Detail from the Foils

[Image 587]

Detail from the corrections / daily values: Change in initial velocity affecting barrel elevation (small form body). The setting of v0 is only made at the reduction body — no foils exist for this. Detail: Horizontal: projectile flight time, from 17.0 (left, cut off) to 22.2 seconds (right). Vertical: flight altitude, from 3,600 m (top) to 9,600 m (bottom), every 200 m. Clearly visible at the bottom is how an extension of the range was subsequently calculated by another hand. A transition zone was calculated twice, and it does not agree perfectly, as elsewhere. There are several individual corrections, which are occasionally also dated (1951). Official HASLER paper; entire table original approximately 80 × 56 cm.

Detail from a large form body: Durata, aircraft flight time from “now” to the point of intersection, for determining the lead and the point of intersection.

Horizontally: flight altitude from 3,300 m to 4,800 m (every 100 m).
Downward: horizontal distance from 9,050 m to 9,750 m (every 50 m). Various corrections at recalculations, with deviations of 0.02 seconds. Around 25.3 sec a boundary is hatched (a very long time for shooting — rather almost illusory). The projectile flight time is approximately 3 to 6 seconds less, i.e. the time from “now” to firing must be subtracted.

[Image 591]

[Page 7]

[Image 589]

For the calculation “flight time,” there is a second (later?) set of 9 foils, described somewhat differently, in which flight times of up to 32 and even up to 40 seconds appear. A step line has been drawn here at 31 seconds. In 30 seconds, an aircraft traveling at 500 km/h covers a distance of 4.2 km.

The flight altitude is called “Altezza” or most often “Quota.” Here, with respect to horizontal distance, computation is carried out only every 100 m.

Images 589 and 591 overlap in the range: altitude 3,800–4,200 m, distance 9,100–9,700 m. Already computed ranges are recalculated, with differences of 0.01 to 0.03 sec; in most cases the new value is minimally larger or equal, rarely smaller.

Images below: In every other column (altitude), every other value (horizontal distance) has been written by another hand — for example in the left image with a completely different numeral “2,” in the right image with different numerals “4,” “5,” and “7.” These may be subsequent interpolations between the actually computed values, or a second person calculated these values to more easily detect errors.

Durata: Horizontal distance 9,450–9,850 m, flight altitude 3,300–4,000 m. [Image 591]

Inclinazione: Horizontal distance 300–700 m, flight altitude 8,800–9,600 m. [Image 593]

How Are the Shapes of the Ballistic Bodies Calculated? How Accurate Are These Calculations?

First, the trajectories of the projectiles must be known — without this information, manufacturing the form bodies cannot even be contemplated. For every firing angle (= elevation angle), one must know when the projectile is where. The deceleration of the projectiles due to air resistance is a particularly hard nut to crack.

Before the trajectories can be calculated, the initial velocity v0 must be known. There are reports from eyewitnesses of German anti-aircraft operations in World War II stating that the velocity of the projectiles was measured primarily mechanically, using the Le Boulengé apparatus. Test personnel would follow the batteries and measure from time to time how much the v0 of individual guns had changed due to barrel erosion. The measuring principle consists of two falling weights being released by two electromagnets as soon as the projectile destroys wires stretched back and forth on a frame. The distance between the two frames corresponds to a flight time of approximately 0.1 sec (at a v0 of 800 m/s, approximately 80 meters). The second falling weight triggers a mechanism that drives a small sharp blade into the first weight, from which the time interval between the two starting times can be read, and from this the initial velocity of the projectile is calculated.

The original descriptions of this apparatus by Le Boulengé (first trials from 1863 onwards) sound cumbersome and complicated; the device was, however, said to be simple to use and thus survived for a long time. Attempts at electronic measurement using vacuum tubes are also likely to have existed during World War II.

[Page 8]

Calculation of Trajectory Curves (Wonder and doubt — with far too little specialist knowledge!)

An antiquarian description of an approximation method for calculating trajectories could be acquired, published in 1937: “Einfaches Verfahren zur graphisch-rechnerischen Bestimmung einer Geschossflugbahn” [Simple Method for the Graphical-Computational Determination of a Projectile Trajectory], by Dr. Erwin Pflanz, Stuttgart (Wehrtechnische Monatshefte, 12th issue, 1937). This work recommends a semi-graphical method that is said to reduce the time for calculating a single projectile trajectory from 10 to 20 hours to approximately 3 to 4 hours, with acceptable accuracy. In small intervals, a continuous manual iteration is performed: old position plus v times t gives the new position; there, due to deceleration, a new velocity — and in each interval the projectile drops somewhat downward due to gravity. Graphically, a pair of compasses is used back and forth between two diagrams, which is meant to make the coupled velocity decrease (horizontal and vertical) more manageable.

The reduction in velocity due to air braking is the crux! How does one know how rapidly the projectile decelerates in air? Particularly around the speed of sound the deceleration becomes especially problematic. The 1937 publication states, without further justification, how great the deceleration from air resistance is at “Standard trajectory No. 5,” depending on the current velocity:

[Page 9]

The subdivision into individual intervals, the choice of consistently unmixed v-powers, whose curves all pass nicely through the origin — all this looks rather arbitrary. The astonishment grows even greater when one notices: already in 1895, Russian Professor Nikolaj Sabudski of the Imperial Academy of Sciences in St. Petersburg (and/or the Mikhail Artillery Academy) chose exactly the same subdivision of velocity, with precisely the same intervals, the same v-powers, and — to all decimal places — the same factors. He only additionally gave three higher velocity intervals, which Dr. Pflanz may have deliberately omitted (so that he could publish without the data becoming usable): v = 419–550 m/s, v = 550–800 m/s, and v = 800–1,000 m/s. Over a period of 40 years, with certainly different ammunition, in different countries, with different knowledge and different measurement or computation methods, nothing is supposed to have changed in the description and computational treatment of air resistance?

That cannot be! One must seriously ask whether the calculated trajectories for individual, specific guns of other countries are at all realistic. Air resistance depends very strongly on the precise shape of the projectiles. If the mathematical approximation for calculating air resistance is incorrect, then the calculated trajectories are also incorrect, which then also makes the form bodies imprecise.

On the trajectory chart for the Swiss 7.5 cm anti-aircraft guns 1938, v0 = 805 m/s: it takes approximately 8 seconds / 5,000 m range (for a steep shot) or approximately 11 seconds / 6,000 m range (for a flat shot) for the projectile to enter the region of the speed of sound (where the uncertainty of air braking will be greatest). Firing was carried out to much greater distances! The projectile begins at Mach 2.3. First closed supersonic wind tunnel in the world at ETH, Prof. Ackeret, 1935/36, 0.4 × 0.4 m, 700 kW, Mach 2.

[Image 85]

The computer 43/50 R with a newly inserted intermediate deck for manual entry of radar data. The man in the middle looks through the telescope and turns the handwheel so that the aircraft remains always in the crosshair (the entire computer rotates in the process). Alternatively, he could look at the follow-up pointers below (coarse and fine range) and keep the “radar angle” pointer always in agreement with the “computer angle” pointer. In this way he copies the radar azimuth angle into the computer. The man on the left does the same with the elevation angle; a missing man at the back maintains the distance (from the rangefinder or radar). In the background: the 7.5 cm anti-aircraft gun 38.

[Page 10]

Dimensioning of the Form Bodies

Once the entirety of the trajectories has been calculated — from the lowest to the highest barrel position — there are no longer major difficulties in writing down all the numerical values for the three large form bodies: flight time, time fuze, and elevation (for the form bodies of the daily corrections with three variables it may look different). Wherever the aircraft is, for whatever value pairs of altitude, distance, elevation, and flight time — there are always nearby points on the calculated trajectories whose values are known. A great deal of searching through innumerable tables will follow, then interpolation — linear or quadratic — and the required value for the barrel elevation or the fuze time can be written down. How the found manual corrections of individual numbers in the HASLER tables are to be explained must remain open.

Practical Implementation of the Gearing

In the GAMMA-HASLER device, form bodies are used extensively for computation. Other mechanical solutions would also have been available. First, three different solutions for mechanical multiplication are shown using arbitrary examples, followed in detail by the automatic servo-follow system of the three handwheels. Why one implementation was preferred over another can today only be assessed with difficulty, given the lack of direct experience.

Multiplication with a form body

Command device GAMMA: The vertical velocity (gears 73, 72) and the flight time (translation, from above) orient the form body 74 vv. Electrical readout with the fine contact 77, which controls the motor 78. The motor output vv gives the altitude difference that the aircraft will still fly until the point of intersection. This value is added to the current aircraft altitude in the differential at the very top center.

Multiplication with lever rods

Command device SPERRY: Two men set (at the very bottom, by handwheel), while watching follow-up pointers 28, 29, the two components of horizontal velocity. From the form body 37 comes the flight time — by means of linkage 32–36, v is multiplied by t, yielding the lead distance s, which is added to the position (two components N/S and E/W). Linkage: similar triangles yield proportionality, which can also be written as a product.

Multiplication with a friction-wheel drive (integrator)

The rotation angle c of the small measuring wheel is proportional to the rotation angle b of the large disk and to the distance a from the center of rotation — hence to the product. If the disk rotates at a constant speed, the output of the small measuring wheel gives the integral of the variable a over time. Figure: Dr. Ing. Alfred Kuhlenkamp, Chief Engineer of the Army Ordnance Office. Ball integrators were also used; see Appendix B.

Historical figure: p. 13/14. Practical implementation of the servo-follow system for a handwheel: electrical aspects: p. 15.

In models 38 and 40, three soldiers each had to enter one of the three input quantities — azimuth angle to the aircraft, elevation angle to the aircraft, and distance — by handwheel into the computer. For the two angles, the aircraft had to be kept neatly in the crosshair of a telescope — the rotation of the handwheel was transferred, for the azimuth angle, directly to the orientation of the entire computer (many hundreds of kilograms!) and thus of both telescopes. If the tracking is not performed smoothly and continuously but jerkily, all subsequent computed quantities will also be jerky. According to still-living eyewitnesses, the distance tracking from the follow-up pointer of the rangefinder (or the fluctuating rangefinder measurement itself) gave rise to the greatest errors.

From model 43 onwards, an automatic servo-follow system was incorporated to control the three handwheels. If the change in angle or distance is uniform, the soldier does not need to do anything; the handwheel remains at rest. Only when the angular velocity changes slowly does manual correction have to be applied. The smoothness and precision of the tracking are thereby increased.

[Page 11]

The following is a description of the concrete implementation for the handwheel for the azimuth angle α (compass direction from the command device to the aircraft):

Copy of a small excerpt from the gear diagram of model 43/50 R. Following figure: At the lower left is handwheel 1 for rotating the entire computer laterally. To the right of it, symbolically, is the tripod standing on the ground. The tripod carries the large base plate, drawn thick and black here, which remains fixed to the ground within the computer. On it, a measuring roller (drawn elsewhere) traces out the actual course of the aircraft to scale (projection onto the map): distance, horizontal velocity, and lead are measured here. The entire computer is rotated around the fixed base plate — formerly with pinion 3, which now serves only for measurement and indication purposes. The power transmission for moving the computer is carried out from model 43/50 R onwards by two large worms drawn below the large base plate.

[Gear diagram, element 689]

Handwheel 1 carries a switch 111 that selects between direct operation and automatic follow-up, by decoupling gear 113. The automatic follow-up position is shown. Electric motor 128 runs at approximately constant speed (where a precise speed was important, such as in the speed meter, centrifugal governors were additionally installed!). Inside the rotating spherical-cap housing, two friction rollers are pressed in, which can be oriented asymmetrically from handwheel 1 via gear 113. One thereby rotates faster than the other. Their rotational motions are fed to a differential gear on both sides, which passes on only their differential motion (subtraction). This difference is applied to a further differential 120, where it is mixed with the position of the handwheel, via gear 119. If handwheel 1 does not change, the rotational motion of the regulating mechanism 129 is passed unchanged to the drive worms; the computer rotates at a constant angular velocity. If the angular velocity of the aircraft keeps increasing, the aircraft lags behind the crosshair; the necessary movement of the handwheel simultaneously corrects both the new angular velocity and the position of the aircraft in the telescope (mixing of position and velocity). The double spring at 132 returns both friction rollers to a symmetrical position as soon as 111 switches to direct manual operation without the follow-up automatic. Differential 120 thereby passes the rotational motion of handwheel 1 unchanged to the worms (the motor is switched off, and the return spring 132 will ease the re-enabling of automatic operation).

The integrator 129 ultimately determines a product: angular velocity (which is generally continuously varying) multiplied by time — giving the azimuth angle increment to the aircraft. A similar follow-up automatic of this kind is also installed for the elevation angle and for the distance.

Historical View Inside the Gears

The Museum of Communication in Bern manages and publishes contemporary photographs from the firm HASLER. The following photograph shows the region of the follow-up automatic for the azimuth handwheel (in the gear diagram on page 12, the functions of this region are depicted symbolically). The photograph most likely shows a command device 43. Depicted is just under one quarter of the lowermost deck.

Upper left (figure on page 14): The large gear is fixed to the ground via the tripod. With the small pinion the entire computer is rotated around the gear. On top of the large gear rests the base plate, on which a measuring roller traces the path of the aircraft (map projection, to scale; distances and angles are measured).

Enclosed housing at left: Protects the sealed spherical-cap integrator. The electric motor for the drive is immediately to the left, outside the image. At the output of the integrator, the gear rotates corresponding to the current angular velocity of the azimuth angle (compass angle to the aircraft). Via two bevel gears, the rotation goes to a differential gear (the inner gears are barely visible in the opening); here the motion of the handwheel (at the very bottom, at the image edge) is added. Together, it then goes to the pinion on the large gear, which rotates the computer.

Not visible, probably further below: the signal path from the handwheel to the integrator, with which the variable output speed is set.

Far right at the bottom: The gear for automatic operation is switched off; it is pure manual operation.

Upper right: Five vertical couplings (two of them cut off), to transfer the signals to the next, higher deck — one of these is the azimuth angle. Immediately at the couplings, three further differential gears are visible to the left, for adding or subtracting quantities.

According to the manual, in automatic operation a typical tracking speed in azimuth of 400 A per second is to be achieved (that is 4 seconds for 90°), in elevation 150 A/sec, in distance 400 m/sec. In pure manual operation this was not quite achievable.

The lowermost deck also contains: two additional handwheels with automatic follow-up (elevation angle, distance), the curved-flight extrapolation, the device heating, the connector sockets of the thick electrical cables with their many contacts (35- to 53-pole): to/from the rangefinder, to the central vehicle (generator) and from there to the guns.

[Page 13]

HASLER AG, Bern, Image HAS19411/01, Museum of Communication, Bern. Photo on glass plate: René Kirchhofer, “ca. 1940–1950.” With kind permission of the Museum of Communication.

Command device Model 43 or 43/50.

[Page 14]

Electrical Equipment

In addition to the mechanical gearing that performs all computations, the entire computer is permeated with numerous electro-mechanical devices: motors, relays, range monitors, limit switches, locking magnets, reset devices, tracer micrometers for electromechanical copying of variables, as well as for lighting, synchros for remote transmission of data from the rangefinder and to the guns. The following is an excerpt from the circuit diagram for model 43, reproducing the electrical equipment of the three handwheels with the follow-up automatic. The distance handwheel is somewhat more complicated, because in addition to the “slant distance” operating mode one can also work with “constant altitude.” Three of the motors shown here have an electric brake; the azimuth angle regulating motor does not. All three regulating motors (distance, elevation angle, azimuth angle) are in this position switched off (pure manual operation).

Model 43

[Image 174]

Sources and Internet Addresses

Technical drawings of the command devices, plans, photographs: From documents at the Aviation and Anti-Aircraft Museum in Dübendorf, contemporary manuals and descriptions for the use of troops or weapons mechanics. Diagram of the follow-up system for the handwheel “azimuth angle”: Gear diagram of model 43/50 R, 1964.

Numerical calculations of the form bodies by HASLER: Negative copies at the Museum of Communication, Bern.

Description of two mechanical analog computers for anti-aircraft use: US device SPERRY (only tested in Switzerland) and Hungarian device GAMMA-JUHASZ (deployed in Switzerland from 1938 and further developed into the GAMMA-HASLER): http://e-collection.library.ethz.ch/eserv/eth:47590/eth-47590-01.pdf. Description of the GAMMA-JUHASZ also at: http://www.analogmuseum.org/library/GAMMA_JUHASZ.pdf

Description of Italian guns and three different fire-control devices “Centrale di tiro” during World War II in Italy (GALA, GAMMA, B.G.S.): http://www.grifoarciere.org/it/difesa-aerea-cnt/esercito-cnt/58-materiali-c-a/155-materiali-contraerei-2. At this address, both for the 90/53 gun and for the GAMMA command device, it is described that Italy used a later fire-control device F/90-B (or “BT”) derived from the GAMMA, made by CONTRAVES, until the mid-1960s. The fact that it was CONTRAVES, and not the more experienced firm HASLER that developed it, that came into play with the mechanical computer seems strange and may be incorrect.

From 1961 to 1997, a newer fire-control device CT 40 was used in Italy by CONTRAVES (single-axis predecessor to the Super-Fledermaus, Switzerland: Fltgt 63). This device worked with a new electronic analog computer. Instead of mechanical form bodies, special computing capacitors were used:

[Page 15]

rotary capacitors whose capacitance depends on the rotation angle in the desired manner. Numerical values are still transmitted via rotating shafts, as in the mechanical computer.

Flak-Kommandogeräte [Anti-Aircraft Command Devices]. Book by Oberstingenieur Dr.-Ing. Alfred Kuhlenkamp, VDI-Verlag Berlin, 1943. The center and right figures of the three multiplication solutions (page 11) are by A. Kuhlenkamp. A copy is held in the ETH Zurich library. Kuhlenkamp also examined captured foreign devices, e.g. SPERRY from the USA.

HASLER 1852–1952. 100 Jahre Fernmeldetechnik und Präzisionsmechanik [100 Years of Telecommunications and Precision Mechanics]. Guggenbühl und Huber-Verlag, Zurich. On pp. 162–167, a contribution on command devices. From it: Image 85, page 10; thanks to the HASLER Foundation.

Appendices — on somewhat more specialized topics

Appendix A: The Numerical HASLER Data for the Form Bodies Do Not Match Well with the Previously Known HASLER Devices. Justification:

a) A form body for “derivation” is unknown in any of the previously known devices — from the GAMMA-JUHASZ through the versions GAMMA-HASLER 40, 43, to the 43/50R.

“Derivation” can mean two things: First, a small lateral deflection of projectiles due to their rotation. Since air resistance acts on the projectile in a more complex way than merely at the center of mass, there are gyroscopic forces, precession forces that deflect the projectile axis from the trajectory axis.

Second, it was common practice in Swiss-French usage to refer to the normal azimuth angle at the gun or command device (compass direction) as “dérive,” in Swiss-Italian usage as “deriva.” This has nothing to do with the gyroscopic derivation and can easily cause confusion. In numerous places where the meaning is crystal clear and where there is no doubt, the designation “dérive” has been found. The outputs of the three variables to the guns on computer 43/50R are labeled “Seite / Dérive,” “Elevation,” “Tempierung / Durée.” If someone were to translate this as “Derivazione, Elevazione, Durata,” confusion would be complete. In the Swiss regulations 1963/64, the Italian term “Deriva” was used. In Image 429 immediately following, the German text also refers to the “Derivationskörper,” making it clear: this concerns the deflection due to gyroscopic forces. Such a form body is completely new!

Derivation body: Upper left corner of the first sheet. [Image 427] Variables: time in quarter-seconds (for the translation…) and altitude (Quota) every 50 m (for the rotation of the body).

Lower right corner of the same sheet. [Image 429]

The first of 12 large sheets. Total time range across all sheets: 0 to 30 sec, altitude range: 0–11,500 m.

[Page 16]

In the known HASLER devices, the derivation is approximately accounted for by forming the long measuring needle — with which the angle to the point of intersection is measured during the simulated, to-scale flight — not straight, but slightly curved. The lateral deviation of the trajectory due to derivation thereby becomes a function of the horizontal distance rather than of time. For a steep shot to great altitudes (small horizontal distance), the lateral deflection in the known HASLER devices will be too small — a deflection as a function of flight time would in fact be more accurate.

Were “derivation” understood in the second sense — as the normal “azimuth angle” — a form body for “derivation” would make no sense. In none of the realized HASLER devices has there ever been a form body in the computational path to the azimuth angle.

In HASLER devices, a large form body has never before been driven by a time variable (only flight altitude and horizontal distance). The smaller form bodies of the daily corrections are driven by projectile flight time and flight altitude — but they only have an output to the azimuth angle for lateral wind. It would make sense that in a newly developed command device the (gyroscopic) derivation is mixed in as a correction to the azimuth angle. However, the numerical values of the small correction bodies for two variables fit on two sheets — the probably newly introduced derivation body, by contrast, requires 12 sheets full of numbers: this was a large body!

Alongside the derivation body, the large form body “Spoletta pirica” (= time fuze = fuze setting of the projectiles) in the HASLER foils is also driven by time (0–30 sec) and flight altitude. While it makes sense that the actual (gyroscopic) derivation increases with time, it is not clear by which time the fuze time should be influenced. This change of variable would be the expression of a more significant modification of the existing HASLER devices, in which the fuze time is determined by the horizontal distance and the flight altitude.

What supports the view that the HASLER foils in question were in fact calculated for a HASLER device to be delivered to Italy:

The collection of the Museum of Communication holds two historical photographs of HASLER devices (opened, lowermost equipment deck), labeled “Italian Command Device,” dated 21 March 1952. Images: HAS21783/01 and HAS21785/01, accessible on the internet.

The computer 43/50R of the Museum of Communication has the three outputs to the guns (azimuth, elevation, time) labeled in Italian — and all three words differ from Swiss-Italian usage:

Museum of Communication (possibly intended for Italy, or Italian trials?)	Swiss usage (regulations, Image 708, 710) — Caution with “Derivation”!
Deriva — Elevazione — Durata	Direzione — Inclinazione — Graduazione

Appendix B: Mechanical Calculation of Trajectories

The trajectories of the projectiles could be calculated and plotted from 1948 onwards on a 4-meter-long “ballistic integraph” of the Artillery. The mechanical integrator by AMSLER, Schaffhausen — with initially four, from 1950 five ball integrators, electrical follow-up for torque amplification, with a curve plotter and a film camera documenting all computed numerical values — is described at: http://dx.doi.org/10.5169/seals-83990. Reference is made there to a simplified “school model” of 1945 (possibly the prototype of 1941/42?).

In the Federal Archives there are firing trials with the 7.5 cm anti-aircraft gun 38, whose results are compared with calculations from 1943 and 1952 on a ballistic integraph.

The integraph mentioned above has been restored; it even bears a name: “Mariandl.” Description of the restoration, with in part better photographs (Bulletin 3/10, Foundation HAM): http://www.armeemuseum.ch/uploads/media/Bulletin_VSAM_d_3-10.pdf

[Page 17]

The device is kept at the Foundation HAM in Thun. Five boxes with contemporary documents are also preserved. It was not possible to find anything more precise here either about how, at that time, the air resistance for various projectiles, various velocities, etc. was known.

Two caveats: The integrating device “Mariandl” is younger than the GAMMA-HASLER command devices. The form bodies of the command devices had long been completed and installed when the calculations with the ballistic integraph only just began.

The air resistance of the projectiles may have been accounted for differently in the manual calculations than in the mechanical computer. In both cases it had to be checked what could actually be done — with the machine or by hand.

The engineers Curti and Dubois, who built the integraph at AMSLER, rely for air resistance on Dupuis, Garnier, Rouzier. Numerical values: from Table B by Garnier in “Mémorial de l’Artillerie Française,” Annexe du 1er fascicule de 1929.

The idea, the hope, the approximation consists in representing the deceleration of the projectile due to air resistance at every point of the trajectory as the product of three separate factors:

Form coefficient (projectile shape)
Ratio (air pressure at flight altitude / air pressure at reference altitude)
An air resistance function F(m) that is supposed to depend only on the current Mach number m.

The current Mach number depends on the altitude, the air temperature, and the current velocity of the projectile.

Great care was taken to account correctly for all atmospheric variables. The descriptions become monosyllabic and uninformative at F(m) — almost nothing is found there. Everything is complicated by the fact that the firing trials section in Thun uses a completely different notation for F(m) than the AMSLER device — there are conversion formulae… (but not within a single trajectory calculation).

p. 34 text, Image 978: One passage was found stating that a different projectile also requires a different F(m) profile!

Rods/carriage at left: Determination of the Mach number from the current altitude, velocity, and air temperature (with square-root function!). Right: Braking deceleration from altitude, air pressure (spiral groove), form coefficient (handwheel below), and F(m). “Herz” [heart-shaped cam] in the center above, approximately 25 cm wide: The shape rotates according to the Mach number (360° + 90°); on the left, the factor F(m) is read off. Here is the great uncertainty regarding air resistance, made concrete in metal!

Four ball integrators (from 1950, a fifth was added). The variables are also transmitted below the tabletop by rotating shafts; there too, trigonometric decomposition of velocity. Center of image: the motor supplies the time value as a rotation angle to all four integrators. The “hodograph” is plotted on paper, and on other papers the trajectory y(x), not shown here. Table width approximately one meter, total length approximately four meters.

Images, source: Firm AMSLER, Schaffhausen; currently at the Foundation HAM, Thun. The mechanical calculation of a single trajectory is said to require approximately 15 minutes — instead of approximately four working days for pure manual calculation (the effort depends strongly on the required accuracy and the step size). Figures according to AMSLER, which has an interest in selling the devices.

[Page 18]

In the “heart” form-body shown in the image above, Mach 1 lies approximately at the 7 o’clock or 8 o’clock position, where the increase becomes particularly pronounced. Estimated freehand, the pickup point at the upper left is approximately Mach 3. The KTA required that the analog computer be capable of computing up to an initial velocity of 2,000 m/s, which is significantly higher than all velocities achievable at the time:

Flab Kan 7.5 cm 1938: v₀ = 805 m/s. 10.5 cm Kan 1935: v₀ = 785 m/s.
20 mm Flab (contemporary): v₀ = 950 m/s (Oerlikon, Hispano-Suiza?)
Bison guns 15.5 cm (from approx. 1988–92): v₀ = 845 m/s with charge 9 and steel shells.

In 1944 (even before the definitive offer), the firm AMSLER explicitly stipulated that the KTA itself would have to supply the binding numerical values for the air-resistance function F(m) at such high velocities in tabular form. The KTA’s further wish that higher apex altitudes of trajectories also be handled — where air pressure drops to as low as 100 mm Hg (apex altitudes above 13,000 m a.s.l.) — is regarded by AMSLER as nearly impracticable for manufacturing reasons and on account of diminishing accuracy. KTA = Kriegstechnische Abteilung (War Technology Department).

There were concrete proposals (1955) for how the analog computer could be converted for rocket projectiles, whose velocity continuously increases while their mass decreases as propellant is consumed.

Principle of the Ball Integrator

Illustrations from the contemporary documentation for “Mariandl.” Today, information on the operating principle of ball integrators is difficult to find on the internet. According to Th. Erismann of AMSLER, the idea goes back to Henry Selby Hele-Shaw in 1885.

Source: Colonel P. Curti, March 1943: Flugbahn-Integraph Curti 1942. Designer: Dr. Fr. Dubois.

The drive roller and the driven roller do not yet uniquely determine the rotation axis of the ball — one degree of freedom remains. A third roller, pivotable at an angle (not shown here), fixes the rotation axis of the ball. The angle of this “guide roller” is set in “Mariandl” beneath the table surface.

The ball rolls such that all three rollers only roll tangentially and do not push sideways, since that would create greater resistance. The angle φ determines the rotational speed of the driven roller.

Source: Mariandl documentation.

The rotations of rollers T (t) and L (φ) are the input variables; the rotation of F is the output. φ must be set by an external linkage such that tan φ equals the function to be integrated.

What is integrated in Mariandl: the two instantaneous velocity components of the projectile (yielding x and y), the change in trajectory direction due to gravitational acceleration (yielding the new direction), and the change in projectile velocity due to air resistance and gravity (yielding the new velocity). All of these integrations are coupled to one another.

[page 19 ends here]

Source: Mariandl documentation.

At the output of the ball integrator, a torque amplification in the form of an “electrical follow-up system” immediately follows, in the same manner as in the command instrument GAMMA-HASLER. Fine electrical contacts (“follow-up mechanism”) control the position of the motor-driven shaft so that it always corresponds to the position of the variable to be amplified. In “Mariandl,” the switching of the currents for the electric motors is already accomplished by electron tubes with control grids. In the command instrument, relays still exclusively switch the high motor currents on and off.

Appendix C: Types and Technical Development of the Command Instrument GAMMA-JUHASZ, Later GAMMA-HASLER

The instrument was continuously further developed. Some of the more notable development stages are listed here. In addition, there were special forms and prototypes that were never produced in quantity (images accessible on the internet in the Museum for Communication attest to this). Examples include radar tracking pointers mounted as “ears” on top, or a telemeter mounted directly on the command instrument.

In the last known regulations of 1963/64, Model 43 is no longer listed as its own version. In the Federal Archives, Models 43/55 and 43/57 also appear, which do not exist in the 1963/64 regulations. In March 1952, two photographs of an open “Italian command instrument” were taken at HASLER (possibly in the construction state); the image can be found in the Museum for Communication, Bern. From 1953 onward: 24 of a total of 83 command instruments were gradually converted in Switzerland to the 43/50 R for radar operation. The type designation indicates the year of development, not the year of introduction into service. In 1957, the first Fledermaus devices appeared (CONTRAVES) with an electronic analog computer, not yet for the Swiss Army. The 7.5 cm Flab Kan 38 was decommissioned in 1967 — the computer loses its purpose.

Year, Type	Development
38	Housing partly round. Production: Hungary, GAMMA-JUHASZ. There are still two v-horizontal knives. Telescopes are mounted at the bottom. Distance carriage: the two flight-direction arms lie side by side.
40	Housing partly round. Production: HASLER, Bern. Now only one v-horizontal knife. Telescopes still mounted at the bottom. Distance handwheel still at the bottom center of the front face. Distance carriage: the two flight-direction arms still lie side by side. Better electrical contacts by HASLER; “HASLER brakes,” new motors. Max. lateral lead, Arm I: 3,400 m, Arm II: 4,400 m. Max. target speed: 300 m/s (previously 110 m/s). Max. distance: 10,000 m (previously 8,500 m).

Year, Type

Development

Housing partly round. Production: Hungary, GAMMA-JUHASZ. There are still two v-horizontal knives. Telescopes are mounted at the bottom. Distance carriage: the two flight-direction arms lie side by side.

Housing partly round. Production: HASLER, Bern. Now only one v-horizontal knife. Telescopes still mounted at the bottom. Distance handwheel still at the bottom center of the front face. Distance carriage: the two flight-direction arms still lie side by side. Better electrical contacts by HASLER; “HASLER brakes,” new motors. Max. lateral lead, Arm I: 3,400 m, Arm II: 4,400 m. Max. target speed: 300 m/s (previously 110 m/s). Max. distance: 10,000 m (previously 8,500 m).

[page 20 ends here]

| 43 | Housing rectangular. Telescopes now mounted at the top; the lateral telescope slightly higher. Two lower handwheels (elevation and azimuth angle) still at the same height. Distance handwheel is now on the right side. Distance carriage: the two flight-direction arms are laterally offset. All three handwheels fitted with new follow-up control. Extrapolation for curved flight (possibly a first?). Higher vertical velocity (dive) up to −200 m/s (instead of −145 m/s). | | 43/50 / 43/50 R | Two different scales: measuring and computing for aircraft / ballistics. Lateral motion still transmitted to the mount via small pinion. Handwheels for azimuth and elevation angle presumably at the same height (no change in the gearing). Two flight-direction motors: fine / fast. Max. lateral lead, Arm I: 4,400 m, Arm II: 5,450 m. Distance carriage locking ?? Radar intermediate deck with new tracking pointers (2 angles on the left, 2 distances on the right). Old tracking pointer for telemeter distance is mounted in the intermediate deck. Azimuth handwheel is slightly higher than elevation handwheel. Lateral motion transmitted via 2 large worm gears; pinion only for readout. Two motors now also for flight time and fuze setting: new micrometers with fine/fast ranges. Both motor outputs added via differential. Two motors also at lead multiplication v_h·t, v_v·t; high-speed distance. New variables in multiplication: lead with loading delay. One fewer differential. Max. lateral lead, Arm I: 4,900 m, Arm II: 6,400 m. Operating-hour counter in the intermediate deck. 8 electromagnets with slip rings, “lateral release,” on the mount? |

Even the most recent Model 43/50 R underwent further developments. Shown below on the left is the electrical circuit from 1952 for scanning the two form-bodies “lead multiplication”: horizontal velocity multiplied by time (time with and without loading delay, respectively). At the top: one electric motor each with brake. Below that, the relays, which keep the high motor current away from the sensitive micrometer contacts. Micrometers immediately below the relays, with K1, center contact, K2. Below that, limit switches. Upper right: the measuring roller, which traces the aircraft’s path on the base plate. Lower right: horizontal-velocity gauge; the motor’s speed is stabilized by a ball centrifugal governor (!).

Lower right: the same detail from the electrical plan, now approximately 1963/64. New fine and fast motors have been added; the micrometer contacts are more complex, i.e., have become two-stage to handle the two speed levels; the limit switches as well. The two-stage arrangement required an entire set of intermediate contacts (six coils) above the micrometers. In both of these versions, the radar intermediate deck is drawn in for receiving radar data.

1952 Federal Archives — Image 616

Ca. 1963/64 AFC Dübendorf — Image 706

[page 21 ends here]

Function of the Circuit Illustrated Above

The form-body rotates and displaces itself autonomously, corresponding to the current flight attitude. A spring-loaded stylus scans the surface of the form-body from a fixed position. The position of the scanning stylus controls, via the electrical micrometer, one (new: two) electric motor(s), whereby the numerical value of the lead from the form-body is copied in a load-bearing manner onto a rotating shaft that is available for further computations elsewhere.

For approximately 25 years, the firm HASLER worked on this mechanical computer, continuously developing it, expanding its measurement ranges, and consistently introducing further innovations.

The plans (gear and electrical plans) for the instruments from Model 40 onward are signed by the Command of Air Force and Anti-Aircraft Troops, or by K+W (Federal Construction Workshop Thun), or by HASLER AG. If significant impetus for further development may also have come from the engineers of K+W, the late mention of this possibility is offered with apology.

Sources and internet addresses: See page 15, before the three appendices A, B, C.

The Flab archaeologist: André Masson, CH-4900 Langenthal.

November 2015 to March 2016