On June 22, a horrific
crash between two D.C. Metro trains killed 9 riders and injured
more than 70 others. Investigators are still trying to ascertain
exactly what failed. Metro Chief John Catoe
noted:
What we found during a special review of the data after the accident was that the track circuit periodically lost its ability to detect trains. This is not an issue that would have been easily detectable to controllers in our operations control center. What the analytical profile showed was that the track circuit would fail to detect a train only for a few seconds and then it appeared to be working again.
Over at DC Metblogs, contributor TONIGM speculates about a possible source for the problem - tin whiskers. As he explains:
When people first started building electric circuits, they used tin metal to solder the interconnections between the copper bits. It wasn’t long before they noticed the tin would get “furry”, growing spiky whiskers as the part was used. These spikes could grow long enough to short out the circuits, and then were so weak that they would break off right after doing so. A smart metallurgist figured out that adding a small amount of lead to the tin alloy stopped this behavior.
Now, lead is something you definitely don't want in your drinking water (although we D.C. residents enjoyed it for several years). In 2000, European regulators launched a jihad against lead in consumer products. The result, according to TONIGM, was that...
...consumers all over the world were getting lead-free electronics, many times without knowing it. Many times the same part number started showing up with lead-free solder, making this trend very hard to track.
So yesterday, I dropped a note to one of my expert friends, who agreed with me that the circuitry in the Metro replacement part, more likely than not, contained lead-free solder. And then, he pointed out the likelihood that the latest Airbus crashes had lead-free solder components in their flight controls.
Could the Metro crash have been causeed by yet another unintended consequence of zealous regulation? We'll see what the investigators determine.
Go here for whole "tin whiskers" post.
Very interesting. Of course one should start from more probable causes (e.g. software bugs, corrosion, faulty designs). I would be very surprised if the industry had started putting out components without finding a replacement for lead.
This effect (tin whiskers) is especially true in space - hence
tin is not allowed in space qualified hardware.
Also happens (in a different mechanism) with cadminum in space -
the cad outgasses then re-deposits, causing shorts as the cad
builds up to a level where it can conduct electricity (cad plated
fasteners are typical in non-space aircraft applications).
Another one where the enviro's might have killed is the parnoia
over the ozone hole and CFC's - which, in "halon" form, are some of
the best fire extinguishing agents available. How many have died or
been maimed due to lack of effective fire supression agents? Halon
is far superior to dry chemical or CO2 hand held extinguishers.
Reference FAA Advisory Circular 20-42C for a good discussion on the
topic.
The great thing about Reason is that it saves me from having to write satires about libertarianism. While in the past, Reason has waited for a couple weeks after I've written a satire to come out with their own serious proposal, this time, they've come out with their own serious proposal before I've even written my satire down. Good job.
What's next? Is someone going to say banning DDT killed millions
and banning asbestos helped the World Trade Center collapse?
Oh, wait...
24Ahead....exactly where do you see a policy proposal in this?
It seems to read like a hypothesis that, if true, would be a
notable example of unintended consequences. I see no specific
proposal, serious or otherwise.
You may wish to read the penultimate sentence.
I've worked in the electronics field for years and I was
initially ready to laugh this story off. But there may be some
truth to it. I really doubt that the issue was tin whiskers, but
many lead-free solder mixtures do behave differently than
traditional lead-containing solders. I'd be surprised if it was
whiskering, but it could have been fractures caused by differences
in ductility and thermal expansion.
Still, I think it's a remote possibility at best. There are many
pieces involved which are much more likely to fail. I'd blame the
software first, then mechanical failure, then electronic hardware
failure.
Erÿk, old friend, you wont get anything useful from Chris Kelly.
He's a slightly more coherent version of the timecube guy.
The fact that the lead got pulled while the part numbers stayed the
same concerns me; it can fool people into skipping tests that
otherwise would be run on the new material as part of validating
it.
Wiping the increasing film of sweat from his brow, LoneWacko
scoffs at the ludicrous theories proposed by online libertarian
publications. He knows what caused the DC metro crash -
illegals.
Pushing his pulled pork sandwich to the side, he started work on
his essay that would blow the whole conspiracy wide open! He knew
what he had to do!
There is still the undiscussed element of possible operator error. Perhaps the driver of the following train anticipated a 'green' signal and overran an unexpected 'red' board. The driver is the last line of defense.
There isn't a real compelling reason to assume whiskers caused
the issue. Lead free parts have been adjusted to an extent to deal
with this anyway.
The new failure mode for popular electronics is high density BGAs
and various lead-less parts anyway (some high lead solutions
actually had a huge tendency to break joints from thermal stress,
ie Nvidia....).
yet another unintended consequence of zealous regulation
Zealously not regulating can't have even worse unintended
consequences?
Reactionary posts like this one just make Reason - and libertarians - look bad.
MAJOR CORRECTION
As it turns out, almost five months ago I did joke about
libertarian support for lead products here,
which ends with:
[This comment was sponsored in part by the Chinese Army
Rehabilitative Labor Center Happy Fun Toy Making Camp/Lead Smelting
People's Cooperative]
It just took Ronald Bailey a few months longer to turn my satire
into a Reason post.
Sounds a little bit conspiracy-theoryish.
I mean, on what possible grounds could one suggest tin whiskers
caused the AirFrance flight crash, when the freaking flight
recorders haven't even been recovered?
(And the Yemeni AirBus that crashed was banned from flying in
France due to poor maintenance, which had nothing to do with tin
whiskers.)
Okay, scratch that. Sounds a lot conspiracy theoryish.
"Driver was texting" is more likely to cause a train crash than tin
whiskers.
"Did Banning Lead Cause the D.C. Metro Crash? "
I've been waiting for something like this to happen for a very long
time. If tin whiskers didn't cause this problem, it will cause
something similar in the future. It's only a question of
time.
jasno,
I've worked in the electronics field for years and I was
initially ready to laugh this story off.
I don't know what part of the electronics field you've worked in,
or what you do, but there's nothing to laugh off here. Those of us
in the aerospace industry were going ape shit about this long
before the law banning leaded solder ever went into effect.
There's a very simple recipe here: remove the trace level of lead
from tin solder, and you'll get tin whiskers. Not if, not
maybe. The only question is when. There is a huge
body of literature out there on this topic.
They really weren't putting trace levels of lead into tin solder
(before the ban) just for the fun of.
This is one of the biggest risks that commercial aviation faces. We
cannot insure the reliability of electronics anymore, because a) we
really can't predict exactly when the tin whiskers are going to
start occuring, and b) the way the global supply chain for
electronics works, there's no way we can tell if a particular
component has lead free solder in it or not.
And it's not just aviation that's at risk here, it's anything and
everything you can imagine that people's lives depend on, that is
also computerized (read: everything that matters, in today's
world).
turkey,
There isn't a real compelling reason to assume whiskers caused
the issue. Lead free parts have been adjusted to an extent to deal
with this anyway.
Thou knowest not what thou babble-ist about.
I am an aerospace engineer. I've spent most of my career
researching advanced materials, and been on more than a few failure
investigation red teams. I'm one of those people who designs
hardware, that lots of other people's lives depend on.
Know this: there is not yet a real fix in sight for the tin whisker
problem. And while I don't know if tin whiskers caused this
problem, it is entirely rational to put it on the board as a
possible root cause. Because tin whiskers happens.
Ron, just in case you weren't sure -- you've hit on a very real
problem here, and the world has not yet seen the consequences of
it. Ignore the clueless few above, who really don't know what
they're talking about.
Though I'm not convinced the consequences of the lead ban were
"unintended" in any particular way. Materials engineers testified
as to the catastrophic problems this would cause, for many months
before the lead ban became law. The environmentalists pushing for
it, frankly didn't care. Any more than Al Gore "cares".
The excuse was that they allowed us "exemptions" from the lead free
solder ban, for "mission critical" applications (which applies to
everything I work on). But everybody knew then, as we all know now,
what I said above. In today's global electronics industry, there's
really no way you can insure that all your circuit board components
have good (leaded) solder in them. Which is why nobody has even
tried to invoke the exemption, because you really can't do
it.
But you all should all be happy to know that the next time you sit
down to a breakfast of printed circuit boards and granola, the
solder you consume will not contain 1% lead.
We had whiskering problems in some parts at an old job of mine.
We had to go to lead-free parts because Europe was a big customer.
Some suppliers had whiskering on their parts, and some didn't.
Guess which ones were cheaper? We went with the non-whiskering
expensive parts, because it was medical equipment.
Most manufacturer's have moved on to non-whiskering solders. But
there's a lot out there that haven't. Government engineering
processes are all about getting signatures signed, not actually
testing the products. So it wouldn't surprise me at all that crappy
parts made it into a DC Metro circuit.
Well, I can say that Ebeneezer beat me to it. I can't argue with
anything he wrote above. Considering the company I work for, I'm
probably a supplier either to his company or to one of their (few)
major competitors -- or all of the above.
We deal with this just about every day. Our aerospace &
military customers commonly ask us for what are called RoHS 5/6
components (everything can be RoHS - Reduction of Hazardous
Substances, a primarily Euro requirement - compliant except for the
use of leaded solder). The reason they do this is for this very
issue. Alternatively we've had a few ask us for devices that are
conformally coated, which entails adding an acrylic (or other
material, but acrylic is particularly common) coating on the entire
device. This is primarily to insulate the device from moisture
& chemicals, but has the added benefit of reducing the negative
impacts of tin whiskers (i.e. it doesn't stop whiskering, but helps
to keep them from breaking off and shorting something out). But
Ebeneezer above is correct in that this is a very big issue and one
that is not adequately understood -- and in the world of global
sourcing, not one that is easy to be sure that you're getting
exactly what you expect in your electronics.
As for the specifics, it sounds like the main concern in the Air
France Airbus 330 crash was due to faulty pitot sensors -- a
mechanical issue. It sounds like the Yemeni Airbus 310 crash is a
much older aircraft still using hydraulic and cable linkages, so I
doubt any critical electronics in that aircraft were new enough to
be lead-free. As for the DC Metro crash, I can conceive of
lead-free solder being an issue, but it certainly wouldn't be near
the top of the list. There are a lot of more-common potential
issues (such as simple user error, or non-whisker-related
electronics malfunction) that suggest it's WAY too early to jump to
a conclusion.
(Note - I am speaking individually as an engineer from my
experience, but I am NOT in any way speaking on behalf of my
employer.)
Most manufacturer's have moved on to non-whiskering
solders.
If you've got such magic stuff, I'd like to know where I can get
it.
What Brad said.
We also had trouble with lead free solders that have higher melting
points, the higher temperatures were causing component failures.
Led to batches of brand new boards coming in bad from the
factory.
That problem has gotten much better, but the tin whiskers issue
really hasn't. Conformal coatings can help, if you can tolerate
them, but we've had component short out that were conformally
coated, when the pins on components get close enough
together.
It's an extremely difficult problem to study. You can't just do
accelerated aging tests, for the same reason that accelerated aging
tests are particularly reliable for polymeric materials. If you
stick a part in an oven, you're changing the dominant chemical
reactions within the material. What happens in the oven at elevated
temperature, versus what happens if you leave the same part at room
temperature, are often just not the same thing.
For you fellow engineers out there who deal with critical
systems and failure analysis, I've always enjoyed reading the
RISKS Forum
(http://www.risks.org). It deals mainly with risks of
automation and other computer systems in systems, especially with
regards to how they interface with humans and human society.
As a chemical engineer, I don't deal directly with the computer
systems most of the time, but the implications for process control
and safety are obvious. If you guys have any other good engineering
failure-analysis related forums and/or organizations, I'd be very
glad to hear of them!
I can understand removing lead from solder for plumbing
applications*, but I have no clue as to why it should not be used
for electrical or elctronic apps.
I'd like to see an explination for this but I'm too lazy to look
for one so I'll sit back until someone gives me one.
*or for that matter from anywhere that has to do with getting into
contact with food or anything that might contact skin in the normal
course of life.
Check out this informative chapter on lead in our lives, from Bill Bryson's excellent A Short History of Nearly Everything.
I worked for several years in the aviation industry and we refused to use lead free solder, even for our European customers. They could either use leaded or go somewhere else...to my knoweldge nobody ever left, it is such a tiny aspect of a contract.
This is part of the environ strategy to devolve our
technological society to a "more human" level. They have been
steadily demonizing elements from the periodic table, starting with
the devil element itself, plutonium, then moving to lead and then
chlorine and the latest one is carbon, the backbone of life itself
on this planet.
It is not a conspiracy - it is a shared vision of the future.
Yeah. They've got chromium and nickel in their sights already.
And when this hits the fan, I'm not sure how we're going to build
anything anymore.
But isn't that just the vision.
Reason on Facebook
Reason on Twitter
Reason on YouTube
Reason RSS
Site comments/questions:
Media Inquiries and Reprint Permissions:
(310) 367-6109
Editorial & Production Offices:
3415 S. Sepulveda Blvd.
Suite 400
Los Angeles, CA 90034
(310) 391-2245