US1404004A - Searchlight window - Google Patents

Description

F. A. BENFORD.

SEARCHLIGHT WINDDW. APPLICATION man JANJ. 192|.

` 1,404,004. Patented Jan. 17, 1922.

Figk

UNITED STATES PTENT OFFICE.

FRANK A. BENFORD, OF SCHENECTADY, NEW YORK, ASSIGNOR TO GENERAL ELECTRIC COMPANY, A CORPORATION OF NEW YORK.

SEARCHLIGHT WINDOW.

Application filed January 7, 1921.

To @ZZ whom it may concern:

Be it known that I, FRANKA. BENFORD, a citizen of the United States, residing at Schenectady, in the county of Schenectady, State of New York, have invented certain new and useful Improvements in Searchlight Windows, of which the following is a n specification.

My invention relates to searchlight windows through which the rays of light which are projected from a parabolic mirror or other optical system parallel to the axis of the mirror or system shall pass into the field of illumination as a beam of parallel rays. Searchlight windows of this general character have heretofore been .made of one pane of plane glass, through which parallel rays pass and issue from the same as a parallel beam; but it was found that such window, even if made of considerable thickness, is liable to crack and break by reason of the high temperature and greatvariations o f temperature to which vit is subjected, and moreparticularly by reason of the sudden and violent gas pressures against its outer surface due to artillery fire in its immediate vicinity, whenA the searchlight is used in warfare. To overcome this serious defect, the plane glass window was built up of strips placed edge to edge. These strips, however, would not sufficiently close the searchlight casing against the entrance of rain or spra to the detriment ofthe source of light. oreover, the glass strips are also liable t0 breakage, although in a less degree, by gas pressure from the outside. In addition thereto, the dull ground edges of the strips intercepted a considerable percentage of light. Y

It has been attempted oflate to overcome these difficulties by employing concavo-convex windows with the convex side on the outside, and these windows actually secured a good closure against the entrance of rain and spray and were safe against breakage; but by reason of the particular curvatures given to these concavo-convex windows, vas will appear further on, the parallel rays which they receive from the optical system are by refraction largely diverted outwardly, so that the benefit of a parallel beam is to this extentlost. This parallax of refractionl is'barelynoticeable near the center of Specification of Letters Patent.

Patented Jan. 17, 1922.

Serial No. 435,722.

refraction of the glass employed, that the two surfaces are practically parallel within the range of incident and outgoing rays and therefore cause the parallel rays of light received, to issue as a practically parallel beam.

In order to fully appreciate the beneficial effects of my invention and the departure of the invention from the prior art, a brief review of the latter becomes necessary, and for this reason the accompanying drawings which constitute a part of this specification display, in addition to the invention also two constructions of searchlight windows, which have heretofore been used and which have been found wanting. Fig. l represents a section of a parabolic searchlight wiinlow heretofore used, with the course of a ray of light through the same indicated; Fig. 2 is a like vie-w of a concentric spherical searchlight window heretofore used, and Fig. 3 is a like view of my improved searchlight window.

In these drawings the thicknesses of the window glasses are somewhat exaggerated in order to trace the course of rays of light through the same with greater comfort; in consequence the angles of divergence necessarily also become exaggerated, but with the assumption of a. thicker glass everything is drawn to scale. The dimensions here under consideration are necessarily small and the partly exaggerated scale is, therefore, justied. 4

In Fig. l, the window glass l() has a concave inner surface ll accurately ground to a true paraboloid, while the outer convex surface deviates from the parabolic form in such manner that if this glass is used as a reflector, with the silvering on the convex face, rays from the focus of the inner face will, after refraction and refiection and repeated refraction, issue from the concave face as a beam of parallel rays. Such para` bolic mirrors are spokenof in the art mirrors corrected for refraction and they serve their purpose fairly well. Such pieces of glass, but with the silvering removed, have heretofore been used as searchlight mirrors, and Fig. V1 represents a'window glass of such description. It will be not-iced that this glass is thicker at the edges than in the center. 4The normal to a point g1 near the edge of the inner face isv designated by R1 and for the outer face at g2, by R2 and these normals intersect the axisat C, and C2 respectively. The axis X of the. inner face parabola is understood to coincide with the axis of the optical system from which the window receives parallel rays. ln the example shown in the drawing Il, makes an angle al ofBOO and R. an angle cof 22o with the axis. One parallel ray L, is indicated in the drawing. YIt strikes the window at or near the edge at thepoint g, at the angle a, (30) to the normal R1. The index vof refraction m of ordinary `lass with reference .to -.air is about 1.52, and the Vray L, enters the window at an angle b, to the normal R1,

being turnedfromits original course to-V ward ythe normal, which, as shown in the.

drawing, is a clockwise diversion; the relfracted ray within the glass is marked by the letter L2. lThis refracted ray strikes' Ythe convex face of the window at g2 and there makes with t-he normal E, an angle a2; it

is there again refracted but since in this caset-rherindex of refraction of air with reference to glass is the refraction 'is away from the normal R12, that is to say, it is counterclockwise, the ray issuing vin the direction L3, making with the normal an angle a2. The

angle of refraction a2 is smallerthan theV angle efrefraction b1, so that the issuing ray L, still forms a clockwise angleV with the axis of the parabola; this angle is, there- .fore, an outwardly divergent angle. The ab-l.

solute value of this. angleof divergence is small, much Smaller than shown in the drawing, Vbut its practical. value is considerable, because it diverts from the central beam a considerable kamount of light into the circumference of the beam andthus reduces the central intensity.

In a. 36 searchlight using a window as.

above described accurate calculations showed a. maximum divergence each Away outward from the axis to be OO, 16', 25, 25, so thatl the actual spread wasV OO, 32, 50, 50, which i is quite a serious matter.

Referring to Fig. 2, the window glass 10 is ...of thespherical type, heretofore used, the

two surfaces being accurately concentric; they have a common center of curvature C1, and all lines and angles are marked withthe light through this window can therefore, `,be

easily traced. The angle of loutward divergence of the ray L3 is, as in F 1, the angle formed by that ray with a line parallel to the axis. `The angle of divergence in thisV case is smaller than in the case of the parabolic-window;V it has been accurately calculated in aBG searchlightrat a point near the edge and was found to have there a tot-al spread of 0, 23', 14, 22, which is still quite a serious matter. f

I It will now be understood that the outward divergence 0f parallel rays of light directed against the concavesurface of winangle between normals toV a pair ofpoints of incidence and exit, respectively, is made prac- Y tically Zero andvconsequentlythere is no divergence. Of course, the concave and. con` vexsurfaces cannot,- as a whole, be parallel,

but to each elementof the concave surface aty the point of vincidence of a parallal ray corresponds an element of the convex surfaceat the point of exit, and byV my invention these conjugatedelements are made practically parallel, and by reference to Fig. 3 the construction bywhich this isfaccomplished is w explained. Y

In Fig..-3 the concave surface ofthe win-A dow- 10 isaccurately spherical, the center of curvature being at C1, and the suitably chosen radius of curvature. is R1.: The convex surface is also Yaccurately spherical, but its center of curvaturev C: is acertain .distance D nearer `the innersurfacefand in this respect Vthe construction diersfrom :those heretofore used and shown in Figs. 1

andvEZ. The as yet undetermined radius'of curvature Vof theV convex face, is R2.

reason fof this construction'the/new window is thinner at the edge than in the center; the ditferencein thickness is small but itis significant, and in order tomakethis Vdifference apparent to the vunaided eye the, window,v

Vglass'lines have` in the `drawing been ex tended beyond the edge of the window.` The thickness of the edge as measured along the radius R2, is denoted byTMandA this thick-` ness can be chosen at willandgis therefore a constant upon which, together with. the lindex ofV refraction andthe arbitrary length of R the values'of D and R2 depend. These are the two critical quantitiesgavhen a rulek for ascertaining theserquantities. has been established; a window in accordance. with my invention-can be constructed. This rule is Y Larrived at by the. following considerations. same letters of reference as the correspondingparts inFig. 1. The course of a ray ofV .parallel may Ll at the point gnbe vdenoted a,

In order that the elements at g1 and g2 be parallel, the normal R2 to g2 must be parallel to R1, and when it is thus drawn it intersects the axis X at the point C2. Consequently S *sima sin l) sin a cos b m cos b The length of R2 (the radius of curvature of the convex surface) is If in this equation the value of D determined above is substituted R2 is found to be COSG.

The quantities D and R2 being thus determined under the condition of parallelisml of the conjugated elements g, and g2, it follows of necessity that the issuing ray L3 is parallelto the incident parallel ray L1; its divergence is therefore zero, and a parallel beam striking the concave surface of the window at g1 issues at the convex surface as a parallel beam.

In the searchlight windows heretofore used the divergence of the issuing rays from parallelism is greatest at and near the edge of the window and decreases toward the center. At the center itself there is no divergence and for some distance from the center the divergence does not'become practically appreciable. By my invention, the divergence of rays at the edge is entirely abolished and for some distance from the edge the divergence does not become practically appreciable. Therefore, in the new searchlight window there is one zero point of divergence (in the center of the window) and one zone of zero divergence (at the edge) andl while in the intermediate space there is, theoretically some divergence of issuing rays, this divergence is as a whole practically not appreciable. The new window, therefore, renders the issuing rays practically parallel.

Since both surfaces of thewindow are truly sphericahthey can be easily ground to scale-whenthe rules for determining the quantities D- and R2 above stated lare observed. Moreover, ,for the guidance of-the manufacturer lof vthis searchlight Iwindow and as a test of the accuracy ofthe construction, the thickness of the central part of the window in terms of the index of refraction and of the arbitrarily chosen thickness at the edge can be calculated from the above formulae. This, however, is a matter within the knowledge of a person skilled in the art and is therefore not dwelt upon.

Accurate tests of the light flux obtained with a 36 searchlight have shown that with the use of a w-indow of my invention the intensity of light projected into the central portion of the beam is about 18% greater than with a parabolic window such as heretofore described.

It is self-evident that my improved searchlight Window is reversible, that is to say, the concave side of it may be placed outwardly, so that the parallel rays from the reector or other optical system may be received at the convex face and will then issue as a parallel beam at the concave face.

Having now fully described my invention, I claim and desire to secure by Letters Patent:

l. A searchlight window comprising a concavo-convex plate of glass having eccentric spheroidal surfaces related to each other to make the parallax of refraction for rays of light parallel to the axis substantially zero.

2. A searchlight window comprising a concavo-convex plate of glass having eccentric spheroidal surfaces, their eccentricity, radii of curvature and thickness of glass near the edge being related to make the parallax of refraction for rays of light that are parallel to the axis substantially zero.

3. A searchlight window comprising a concavo-convex glass having on its opposed Surfaces a pair of conjugated substantially parallel elements for each ray of light parallel to the axis incident upon either one of the surfaces. Y

4. A searchlight window comprising a concavo-convex glass having opposed eccentric spheroidal surfaces. each surface having an element that is substantially parallel to a corresponding element on the other surface for each incident ray that is parallel to the axis. y

5. A searchlight window comprising a concave-convex plate of glass having opposed eccentric spherical surfaces, so related that the normal to any point of incidence of a ray of lightsubstantially parallel to the axis is parallel to the normal at the point of emergence.

normal at thefpont (if-emergence and the v l normal 'to any pointv ofrncdence between the edge and` center is'substantially parallel to the corresponding point of emergence.

In Witness whereof, I have hereunto Vsei; my hand this 6th day.r of January, 1921.'

RAK A. QBENFORD.

Images