SEVERAL YEARS ago Michael E. Meichle, who was then a student at Skyline High School in Idaho Falls and is now a freshman at the University of Montana, set out on an unusual line of experimentation: smashing glass. He broke old window glass, new window glass and occasionally lamp bulbs. In 30 months of experimenting he destroyed some 200 pounds of glass. In addition he made more than 4,000 high-speed photographs of glass-smashing and compiled a report of the experiments that won a top place for him at the 21st International Science Fair. The experiments demonstrated that glass fractures in predictable patterns and that the strength of the material can be altered by fairly simple procedures. Meichle discusses his experiments as follows:

Three years ago a small telescope mirror that I had been grinding and polishing for nearly a year accidentally slipped from my hands and landed on the concrete floor. Instead of shattering as I expected, the glass simply cracked. Moreover, it continued to crack for several days. I could see the fracture advancing steadily toward the edge and thought the glass would eventually fall apart in my hands. Instead the growth rate of the crack decreased as the internal stresses gradually approached equilibrium. The disk remained intact. It was useless as a telescope objective, of course, but it was not a total loss. It introduced me to an engrossing new hobby: making experiments that show the basic fracture patterns of glass and demonstrate that various surface coatings influence the resistance of glass to fracture by impact.

"I started the project by making a number of assumptions about fracturing and then testing each assumption with an experiment. For example, it seemed reasonable to suppose the strength of glass would depend in part on its chemical composition. In other words, the strength of soda-lime glass might reasonably differ from that of lead glass or borosilicate glass. Similarly, it seemed reasonable to suppose the size and thickness of a specimen would influence its strength, as would the condition of its surfaces. A small, scratched specimen might tend to break more easily than a large specimen with unmarred surfaces.

"One important assumption was made in the interest of minimizing the cost of the project. It seemed reasonable to suppose glasses of all kinds might fracture the same way even though each kind differed in its resistance to fracture. For example, a piece of relatively costly borosilicate glass might be more resistant to fracture than a comparable specimen of inexpensive soda-lime glass, but both pieces would fracture similarly when subjected to an appropriate impact. On this assumption most experiments were made with specimens of window glass. The experiments demonstrated that window glass fractures in predictable patterns. Subsequent experiments proved that costly glasses fracture the same way.

"Most specimens consisted of new window glass cut into squares that varied in size from one inch to four inches. Distributors of window glass routinely fill orders for panes of specified dimensions by cutting narrow strips from sheets of standard size. The strips are discarded. When I explained my project to our local hardware dealer, he invited me to help myself from his scrap barrel. I chose strips of appropriate width and sliced them into squares with a glass cutter of the wheel type.

"I learned to choose the glass with care. The strength of the material is altered by the condition of the surface. Old glass tends to be unpredictably weaker than new material, particularly if it has been exposed to weather. The strength of new glass is reduced by surface defects, including minute scratches that may be almost invisible.

"Window glass is available in at least two thicknesses: standard and double weight. When making tests, the specimens must be grouped according to thickness. Indeed, I segregate all specimens that are cut from the same sheet and test them in sequence so that account can be taken of manufacturing variations in the quality of the product that occasionally arise from differences in composition, annealing and handling.

"I assumed that the apparent strength of a sheet of glass would be influenced by its supporting structure. For example, a sheet that is supported by a solid anvil and receives a fracturing blow in the center might break differently from one clamped at the edges so that it can bend and thus yield in the direction of the force. To check this assumption I made a simple test fixture [see Figures 1 thru 4]. It consisted of a slab of plate glass six inches square and 3/8 inch thick and two strips of the same glass two inches wide and six inches long. The square slab could be supported over a hole in the top of a strong box that contained a xenon flash lamp for making high-speed photographs. A specimen to be broken could be put directly on the slab for support over its full area. Alternatively, the two strips could be put on the slab and spaced at any distance, so that the specimen could be placed across the strips, like the span of a bridge, for support on two sides but not in the middle. A force directed downward at the unsupported middle would cause the specimen to bend more or less depending on the spacing of the slabs. Non-uniform bending can be induced by placing the strips at an acute angle.

"All specimens were broken by impact. The scheme used most frequently for developing reproducible impacts entailed dropping a steel ball of known mass onto the glass. I assumed that the minimum force of impact required to break the specimen gave a measure of the relative strength of the glass. Because impact varies with the height from which the ball is dropped, the relative strength of a specimen can be recorded as the minimum distance through which a ball of known mass must fall to break the glass.

"Several devices were built for dropping the ball. The simplest scheme worked best. It is a wood clothespin of the coiled-spring type fitted with a peep sight made of wire. The ball is clamped in the jaws of the clothespin. To break a specimen I hold the assembly above the glass at the desired height, take aim and release the ball by squeezing the handles of the pin. As an aid m aiming the ball I place a rectangular grid of ruled lines behind the fixture that supports the specimens. The force of impact varies with both the mass of the ball and the height from which it is dropped. I used ball bearings in a range of diameters from 1/4 to 15/16 inch. Most specimens were broken with 3/4-inch and 15/16-inch balls.

"A few specimens were broken by placing a steel rod vertically against the glass and striking the top of the rod with a hammer. This scheme was particularly handy for making high-speed photographs of fractures, although the force of the impact could not be closely controlled. The rod and the head of the hammer are connected to the trigger circuit of an electronic flash lamp [see illustration lower left]. They function as an electric switch that closes when the head of the hammer hits the rod. The lamp can be triggered in advance of impact by fastening an extension of aluminum foil to either the top of the rod or the head of the hammer. The interval of time between the flash and the impact varies with the length of the foil extension. By adjusting the length of the foil one can photograph the fracture at various stages of development.

"I also built an apparatus for breaking lamp bulbs. It consists of a weighted arm hinged to a horizontal base [see illustration at lower right]. The arm swings in the vertical plane and carries a steel rod at its free end. When the arm falls, the end of the rod strikes the bulb perpendicularly to its upper surface. The bulb is supported horizontally by a socket. The lower surface of the glass rests on a block of wood that is solidly mounted to the base.

"It is possible to determine the final pattern of fracture by reassembling the fragments of a broken specimen, but the operation is tedious and time-consuming. My observations were made with high speed photography. I used a 35-millimeter camera and a $15 electronic flash lamp.

"Although the camera has an electric switch for synchronizing the flash with the operation of the shutter, I did not use it. Instead I set up experiments m a room that can be darkened. After the camera has been focused on the specimen I darken the room, open the shutter for a time exposure and trigger the flash by one or another of three devices that are actuated by the experimental apparatus. The shutter is then closed by hand. Light reaching the film is provided solely by the flash. The exposure time is determined by the duration of the flash. The flash of my lamp persists for about .0005 second. A flash duration of .001 second or less is adequate for photographing most experimental events, including the fracturing of a lamp bulb.

"One of my three schemes for triggering the flash consisted in wiring the flash unit to a steel rod and the head of a hammer. In another scheme the trigger circuit is connected to a pair of narrow strips of aluminum foil separated by about an eighth of an inch. The strips are placed in the path of the falling ball. The ball bends the strips into contact, thus closing the circuit that flashes the lamp. A thin strip of wood cemented between the foils at one end supports the structure.

"Switches of this general kind can be constructed in various obvious ways. All such switches close the triggering circuit in advance of the moment of impact. Electronic flash lamps do not reach full brilliance at the instant the triggering circuit is closed. The characteristic time lag that separates the two events can be determined experimentally.

"Most of my photographs were made with a third device, which triggers the flash after impact. The apparatus is known as a "sound slave." Noise made by the impact triggers the lamp.

"Sound slaves can be bought for $20 and up from dealers in photographic supplies. I made one that is described by Ray E. Pafenberg in the January 1965 issue of Popular Electronics. It consists essentially of a microphone that picks up the sound of the impact. The resulting electrical pulse, after amplification, operates a fast electronic switch. The switch closes the triggering circuit of the flash lamp.

"The interval of time between the impact and the flash varies with the distance between the microphone and the specimen. Sound waves travel at a velocity of about 1,100 feet per second This delay must be added to the characteristic interval between the instant when the triggering circuit is closed an the moment when the flash develop maximum intensity. Care must be taken to avoid spurious or unwanted flashes. "The microphone should be placed on a soft pad for insulation against vibration. I learned to walk softly and breathe quietly. Fortunately flash units require a few seconds to recharge after delivering a flash. I take advantage of this interval to perform noisy operations that would otherwise trigger unwanted flashes.

"For example, after placing the specimen on its fixture and loading the clothespin with a ball I turn off the room lights. I then move to the position from which I shall release the ball. Next I turn on the power and test the flash lamp by snapping my fingers. As the flash unit recharges I open the shutter of the camera and wait for the 'ready' light to appear. The ball is released. If all goes well, the apparatus makes a photograph of the fracture automatically. I then close the shutter, turn off the photoflash apparatus, turn on the room light and prepare for the next experiment.

"From what height should a ball of known mass be dropped to break a given specimen? How far should the microphone be placed from the specimen to photograph the event? The answers must be determined by experiment. The ball is dropped from increasing heights until the specimen breaks. Another specimen of identical size is then placed on the fixture and subjected to the same final impact. It may fail to break. (The first specimen may have been marred and thus weakened by repeated impacts.) The height of the drop is increased and the ball is dropped again. Eventually the experimenter finds the minimum height for breaking most identical specimens. Enough specimens are then broken to compile a statistically valid result.

"Little can be learned by fracturing a sheet of glass unless the event is photographed. At first it appeared as if the cost of photographic supplies might exceed the limit of my pocketbook. I found several ways to reduce the cost. For example, I used 35-millimeter film because it costs less than larger sizes. Moreover it can be bought in bulk rolls and hand-loaded into cassettes at a saving of almost 75 percent over the cost of preloaded cassettes. In addition 35-millimeter film in bulk is available on the surplus market at a fraction of its retail price. The surplus material is outdated and inferior in quality to new stock, but it is adequate for testing apparatus and making nonessential records that will not be printed.

"I also made substantial savings by developing my negatives at home. It is not necessary to master the art of making prints or to invest in a projection printer. I judged the results of experiments by examining the negatives directly. Prints were made of relatively few exposures, primarily for use in my final report. Learning to read the negatives has a secondary advantage. The results of an experiment can be examined as soon as the wet film comes out of the fixing solution, usually within less than 15 minutes.

"Although fracture patterns of many kinds were recorded during the course of the experiments, four basic types were served in every experiment, either singly or in combination, and their causes were identified. The basic patterns of fracture can be produced at will by establishing appropriate experimental conditions. For want of a better term I refer to the basic patterns as fracture zones, primarily because I originally supposed that the basic patterns were confined to limited areas of the glass. The assumption was subsequently disproved, but I retained the name nonetheless.

"The experiments demonstrated that a sharp blow by an object of low mass and high velocity will create a cone shaped fracture at the point of impact on the surface of a sheet of glass that is rigidly supported. I refer to the break as a Zone 1 fracture. It is the first fracture to appear after impact and seems to be caused by shear stress.

"To confirm this assumption I made a model of clear plastic that represented the cross section of a glass sheet and loaded it at a certain point to simulate an impact. When the loaded model was examined with a polariscope, shear stress could be seen in the pattern of a Zone 1 fracture. (Incidentally, a small pellet at high velocity that strikes a thick sheet of glass, such as an unsupported slab of 1/4-inch plate glass, can transmit a shock wave through the material from the Zone 1 fracture that detaches a smooth cone of glass from the rear surface. The detached cone leaves a clean tapered hole through the slab that diverges from front to rear. I am told that such holes frequently appeared in the windshields of automobiles in the days of gravel roads before the introduction of shatterproof glass.)

"Zone 2 fractures consist of a radial pattern of cracks that spread from the point of impact. The pattern originates in the Zone 1 fracture and appears to be caused by a shock wave that radiates from the impact. I recorded the shock wave by making high-speed color photographs of breaking specimens through crossed sheets of Polaroid. One polarizing sheet was placed in front of the flash lamp so as to direct polarized light upward through the specimen. The second sheet was placed in front of the camera lens and was rotated to the position where minimum light is transmitted by the unstressed specimen. The apparatus therefore acts as a polariscope.

"Fractures of the Zone 3 type appear when the glass is not supported evenly or when the impact is applied at an angle with respect to the surface of the glass. As with all fractures, the glass fails under tension. Tension may develop when an area of glass is pushed in the direction of the impact or bent in the direction of least support. Stress patterns that accompany Zone 3 fractures were also photographed in polarized light.

"Fractures of the Zone 4 type appear when glass is caused to bend substantially under impact. The deformation creates tension in the lower surface. The fracture is explosive when the force exceeds the strength of the material. The fractures radiate from the point of impact in the direction of least support. For example, if opposite edges of a square specimen rest on a pair of parallel supports and impact is applied at the center, the fractures radiate in the form of an hourglass. Zone 4 usually develops as the final pattern of a complex fracture.

"The four zones appear either singly or in combination when soda-lime glass is broken by impact. Their distribution can be predicted if the physical conditions under which the break occurs are known. By taking zone characteristics into account one can predict how a glass object of complex shape will break. For example, a Zone 4 fracture can be expected in an area of the object placed in tension by a bend. Similarly, one would expect to find Zone 3 fractures in a region of glass that is stretched by impact.

"I found it interesting to predict the patterns into which a lamp bulb would fracture and then check the accuracy of my guess by experiment. First I examined the shape of the bulb and took into account its supports as well as the point where the impact would be applied. I then broke the bulb and several others so that fracture patterns could be photographed at various stages of development. I would have preferred to break a single bulb and photograph the developing fracture with a high-speed motion-picture camera, but this equipment was beyond my means. I satisfied myself by experiment that similar lamps fracture the same way, as depicted in the four accompanying photographs [above, left and right] with superposed numerals that indicate fracture zones by type. Impact was applied to the upper surface of the bulb by the falling rod. The glass did not bend at the top of the curve. A Zone 1 fracture developed at the point of impact. Primary shock at this point initiated Zone 2 fracturing that radiated as needle points from the Zone 1 fracture. The shock of inrushing air created a force that tended to tear the lamp from its socket and stretched the triangular area at the top of the bulb near the base. As a consequence needle-shaped fractures of the Zone 3 type developed in that area. As the action continued Zone 2 fractures were stressed inwardly by the moving air, putting the sides of the bulb in tension. The bending caused Zone 4 fractures to develop randomly in the sides. Finally, Zone 2 fractures broke off and were pulled completely inside the bulb. The zones continued fracturing and the fragments separated. The bulb collapsed.

"During the course of the experiments I observed that the resistance of sheet glass to fracture by impact increases with area. When a series of specimens are broken under identical conditions, including equal impacts, the size of the fractures varies inversely with the area of the glass [see illustration at left]. Surface coatings also influence the resistance of glass to shock. A soiled or wet window of soda-lime glass will fracture more readily than a clean window. Conversely, window glass that is coated with a film of egg albumin or sodium silicate and then dried is substantially more resistant to fracture by impact than clean glass. My investigation of the influence of such coatings was limited. The accompanying chart [below] displays the results of a few experiments.

Bibliography

THE PROPERTIES OF GLASS. George W. Morey. Reinhold Publishing Corporation 1954.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.