headlamp filaments after severe auto crash
	headlamp filaments with melted glass shards
new taillamp filaments

Were the Lights On? Lamp Analysis after Vehicle Accidents

Many nighttime accidents result simply because one driver did not see the other. Not surprisingly, in the aftermath of an accident, a question as simple as, "Were the headlights on or off?" will often yield completely different answers, not only from the drivers involved in the accident, but from the eyewitnesses too. Thus, it is important to be able to answer the question from an unbiased, objective, scientific perspective. For that matter, it is also important to answer the same question about the turn signals, brake lights, and running lights, as they also provide important information about the drivers' actions just before the accident.

Some of you may have seen vehicle lamp examinations featured this year on an episode of CSI, the popular new television show about crime scene investigation set in Las Vegas, Nevada. It probably comes as no surprise to our long-time readers that vehicle lamp analysis is an activity we've been involved with for many years. It turns out that in many instances, if you know what to look for, the questions can be answered quite conclusively.

After-the-fact examination of just the vehicles themselves, without a detailed analysis of the lamp filaments, may not be sufficient to determine whether the lights were on at the time of impact. That is, controls may have been bumped on or off during the accident, wiring that supplies power to the lights may have been severed, filaments may have been burned out prior to the accident.

So how does one go about determining if the lamps were on? Let's start with a short lesson about lamps.

As most of you may already know, the light source in a bulb or sealed-beam lamp is a coiled filament made of tungsten wire. (It looks a bit like a smaller, fancier version of the spring in a ball-point pen). When electrical current is supplied, the filament is heated to a temperature above 4,000F and it glows… Brightly! Now, in air, a tungsten filament would react with oxygen and burn out in a matter of seconds. So to prevent oxidation, the filament of a conventional lamp is surrounded by a glass bulb, and the bulb is either under a vacuum or filled with an inert gas.

A halogen bulb has a quartz or other special glass envelope, and it's filled with an inert gas plus a small amount of a halogen gas (iodine or bromine). What does this accomplish? Plenty. In a conventional bulb, the glass is gradually blackened as the result of deposits of tungsten evaporated from the filament. In a halogen bulb, the tungsten vapor reacts with the halogen gas to form a tungsten halide gas.

The halide decomposes at a high temperature and returns the tungsten to the filament. Presto! Your tired, old filament is continually rehabilitated.

 

	In both types of lamp, the filament ends are attached to support posts that feed current in from the contacts at the base of the bulb. During a severe impact (like in a car crash!) the unsupported center of the filament tries to keep going relative to the posts, so it bends and gets twanged like a guitar string. (You can demonstrate this with a Slinky.) And it’s the filament’s response to that twanging that gives us an important clue as to whether it was on or off. How so? Well, as we already said, when a filament is on, its temperature is above 4,000F, and at that temperature, the tungsten becomes very ductile. That is, it will stretch and deform a huge amount, in contrast to its brittle behavior at room-temperature. A clean break of the filament means it was probably cold at impact. (see photo at left). A deformed filament with looping or twisting of the wire means it was hot at the time of impact (see photo at top of page).
As you might expect, the degree of deformation is a function of the impact force (among other things.) Years ago, we ran a series of laboratory tests on lamp filaments where we exposed them to varying conditions of impact and examined their resulting behavior. The figures at the right show the filaments in a headlamp both before (top) and after (bottom) an impact. Notice in the case of this fairly mild impact, the filament is much less distorted than in the case of the filament removed from a vehicle after a severe car crash (top of page). OK then, how about a turn signal lamp, which is alternately on and off? Interestingly, in a typical turn signal, the duration of the off cycle is short enough that the filament doesn’t cool below the ductile-brittle transition temperature of tungsten and thus, its behavior remains ductile. Therefore, even though the filament may be momentarily off at the time of impact, it can still yield and bow in a ductile manner.

Of course, like everything else in this line of work, things aren’t always this simple. If the impact is not severe enough either to deform or break the filament of an intact lamp, then it is not possible to determine whether the filament was on or off. Also, a lamp filament located more than about five feet from the impact point generally will not deform even if it is on. Another complication is that a lamp may have been burned out prior to the accident and not replaced.

How about cases where the bulb glass is shattered by the impact? It turns out that this is oftentimes a real plus to the investigation. First, if the bulb was on, then the air that enters causes the filament to burn up and produce tungsten trioxide, which is a yellowish powder that deposits itself on other cooler surfaces.

Second, when a filament is on, its operating temperature is above the fusion temperature of bulb glass; thus, if the impact shatters the glass and sprays the hot filament with glass shards, these will melt and fuse to the filament, clearly showing that the filament was hot during the impact. The photo below shows bulb glass fragments which fused to one of the filaments of a sealed beam headlamp that was broken in a lab experiment. Notice the other filament does not contain any fused glass fragments, clearly showing that it was off at the time of the test.

Of course, another complication is that oftentimes most of the filament may have been lost at the accident scene. However, even the remaining broken stubs offer a clue regarding the lamp condition at impact.

Examination of the end surfaces with a scanning electron microscope (SEM) will reveal the fracture mode. A brittle fracture (a straight or stepped end) indicates the filament was cold, and a ductile fracture (a stretched end) indicates the filament was hot.

Ductile fracture occurs by the initiation and growth of microscopic voids. Voids usually begin at microscopic impurity particles present in commercial materials. After the voids grow and link up, the fracture surface consists of dimples. (The dimples sort of resemble the overlapping pattern of scoop marks you see in your ice cream carton after you’ve dished out a few helpings.)

All in all, lamp analysis is a fascinating subject, one in which a component weighing just a fraction of an ounce can help “shed some light” on the question of what the driver of a multiton vehicle was doing just before an impact.

Lamp Filaments

As the preceding article discussed, analyzing lamp filaments after an automobile accident involves a fair amount of science. It turns out the filaments themselves involve a whole lot of science and a whole lot of technology. We thought our readers might be interested in learning a little more about the science and technology of making lamp filaments.

Let’s begin by stating that other than for a few special exceptions called glassy metals, all metals are crystalline and usually consist of many crystals. The boundaries between the crystals are disordered, and thus have more energy than the adjacent crystals. At elevated temperatures, given enough time, the crystals will grow larger and larger in order to minimize the area of the boundaries and the total energy. If lamp filaments were made of pure tungsten, this process would only stop when the entire cross section of the filament had become one large crystal. OK, so what?

The “so what” is that a single, large crystal structure would make a lousy lamp filament because the wire would deform easily as the result of sliding between the adjacent grain boundaries (just like sliding the cards in a deck over each other). Continued sliding of the grain boundaries would eventually cause adjacent coils of the wire to touch, which in turn would result in less resistance, increased current, increased heat, and self-destruction of the filament.

So, if pure tungsten would behave this way, then you've probably guessed the tungsten used in lamp filaments must not be completely pure. You're right! It is "doped" with certain impurities that modify its behavior from that of its pure state.

The basic technology has been around for quite some time. In fact, more than seventy years ago wire manufacturers learned a very clever set of tricks to make much better "doped non-sag" wire with elongated, interlocked crystals.

The technology used to fabricate tungsten lamp filaments is unusual because it does not involve melting, which would prevent use of these tricks. Instead, here's what's done:

The tungsten wire used to make filaments begins as tungsten oxide powder. The oxide powder is chemically converted (reduced) to metal powder, and the tungsten powder is doped with small amounts of three impurity elements -- aluminum, potassium, and silicon. Potassium is the critical element, since during processing some of it gets trapped and forms submicroscopic vapor filled bubbles that pin the crystal boundaries and stop the grain growth process.

The powder particles are heated to bond them together into a solid billet. Then, the billet is extruded and drawn through dies to make wire. The tungsten wire is wound around a metal wire mandrel to make a coil, and then the mandrel is dissolved.

The end result of all this is a so-called "non-sag" tungsten wire. The wire only contains about 0.05% dopants, which, you might say, is a good example of how to do something with almost nothing.

IMS-ASM Award Winner

-- Avalokitesvara Statue

In our Spring 2001 newsletter, we told of a recent assignment in which we were asked to determine whether a Chinese cast-iron statue in the possession of a Portland family for nearly a century was of ancient origin or merely a century-old souvenir. As some of our readers may recall, we were able to establish by metallurgical and chemical analysis, in combination with our knowledge of ancient metallurgy practices, that the statue was indeed of ancient origin.

We thought the story was intriguing enough to be worthy of entry into the International Metallographic Society's (IMS) International Metallographic Contest, which is held in conjunction with their annual convention. The contest features the best work of metallographers and microstructural analysts from around the world, as they compete for prizes in eleven classes, and for the prestigious Best in Show Jacquet-Lucas Award, jointly conferred by IMS and ASM International.

This year's convention, the 34th annual event, was held in Indianapolis, Indiana the first week of November, and we were thrilled to learn that not only had our entry won its class, but it had also won the overall, Best in Show Jacquet-Lucas Award. The award consists of a gold medal, an engraved certificate, and $3,000.

For those of you who may have missed the story in our earlier newsletter and are interested in the details of the project, you can find a copy of the newsletter on our web site, www.mei.com/2001/spring01.htm. Or, if you prefer, contact us and we'll be happy to send you a copy of the newsletter.

MEI-C People

Mark Habel and his wife, Jennifer became the proud parents of a baby boy on 9 November 2001 at 11:42 pm. Caleb Hugh Habel weighed 6 lbs, 13 oz, and was 20 1/2 inches long. Congratulations, Mark and Jennifer!

In other wonderful news, Antoni Leander Sharp, born 15 weeks premature to Allen and Valeria Sharp on 23 September 2001, finally came home from the hospital on 31 December, just in time for the new year.

After 14 weeks in the hospital and a starting weight of only 1 lb, 5 oz, Antoni has filled out to a relatively chunky 6 lbs, 12.2 oz and is doing great.

 

23 Sep 2001	31 Dec 2001