Monday, December 1, 2014

The phases of ammonium nitrate and the Takata/Honda airbag recall

I have been remiss in not posting on this November 20 NYT article flagged by an anonymous commenter (thanks!), which explains the detailed chemistry on why the Takata airbag recall may be happening: 2001 Takata had switched to an alternative formula, ammonium nitrate, and started sending the airbags to automakers, including Honda. That compound, according to experts, is highly sensitive to temperature changes and moisture, and it breaks down over time. And when it breaks down, it can combust violently, experts say. 
“It shouldn’t be used in airbags,” said Paul Worsey, an expert in explosives engineering at the Missouri University of Science and Technology. The compound, he said, is more suitable for large demolitions in mining and construction. “But it’s cheap, unbelievably cheap,” he added. 
More than a decade later, that compound is at the center of a safety crisis involving Takata and its airbags. More than 14 million vehicles with the Takata-made airbags have been recalled worldwide over concern that they can explode violently when they deploy in an accident, sending metal debris flying into the cabin. At least five deaths have been linked to the defective airbags. 
On Thursday, Takata’s decision to change the propellant is expected to be among the lines of questioning before the Senate Commerce Committee, which is investigating Takata’s defective airbags. Alby Berman, a spokesman for Takata, said the switch to an ammonium-nitrate-based propellant was not driven by cost considerations. Instead, the company’s engineers determined that the compound produced gas more efficiently with fewer emissions. “This breakthrough allowed us to make the smallest, lightest inflaters available, as well as significantly improve manufacturing safety,” Mr.Berman said. 
Two former Takata engineers said they and other employees had concerns over switching to such a risky compound. “It’s a basic design flaw that predisposes this propellant to break apart, and therefore risk catastrophic failure in an inflater,” said Mark Lillie, a former senior engineer with Takata at its propellant plant in Moses Lake, Wash. Mr. Lillie recently shared his concerns with Senate staff members.  
“It was a question that came up: Ammonium nitrate propellant, won’t that blow up?” said Michael Britton, a chemical engineer who worked with Mr. Lillie at the Moses Lake plant. “The answer was, not if it stays in the right phase.”...
If you were surprised to read about phases in the New York Times, prepare to get more surprised:
...But tetrazole, which is produced in limited quantities and can be expensive, started to squeeze margins at Takata, especially as the airbag market became more competitive, Mr. Lillie said. 
By 1999, Takata researchers in Michigan, pressured by executives, developed a propellant based on ammonium nitrate, he said. But the engineering team in the Moses Lake plant raised objections to basing a propellant on such a risky compound. To bolster its case, the team pointed to explosives manuals warning that the compound “tended to disintegrate on storage under widely varying temperature conditions” with “irregular ballistic” consequences, Mr. Lillie said. 
Ammonium nitrate cycles through five solid states. As the vehicle goes from receiving the heat of sunshine to the cold overnight, the temperature swing is large enough for the ammonium nitrate to change from one phase to another, experts say. Ammonium nitrate also absorbs moisture from the atmosphere readily. Those two things together make ammonium nitrate tablets prone to damage, experts say. 
A focus in the mushrooming recalls has been that the airbags are more susceptible to malfunction in high humidity areas. “Speaking generally, ammonium nitrate can be unstable. Its crystal structure can change according to temperature,” said Katsumi Kato, an assistant professor in safety engineering at Japan’s Fukuoka University. “It changes the burn rate. It leads to various malfunctions.”... 
...Still, at Takata, the answer at the time was to try to stabilize the ammonium nitrate to try to mitigate those cycling effects, but there are limits to just how far ammonium nitrate can be stabilized, said Mr. Worsey, the explosives expert. 
It looks as if the phase change from β-rhombic to α-rhombic happens right at 32°C, which is pretty warm, but still, right in the middle of summer temperatures in the American South. Yikes -- looks like some Takata engineers fought the good fight in this, but the bosses (and The Almighty Dollar) overruled them. Oops.

(The Senate hearing referenced in the article happened, and other than angry people and embarrassed Takata executives, there was no real news that happened.) 


  1. The temperature in cars (and thus of the ammonium nitrate) can swing much more wildly than outside of the vehicle. Mid-summer, temperatures can reach 50-60 Celsius in a closed car, and mid-winter, in northern climes, can be as cold as -30 Celsius mid-winter overnight.That's cold enough for the tetragonal V phase, so two phase changes can occur (tetragonal V <--> β-rhombic <--> α-rhombic).

    1. I saw a phases table for A. Nitrate that associates volumetric changes as swings in temperature drive the 3 phases ref'd by 'pbulsink'. In a perfectly sealed canister, expect minor rise & fall of internal pressures with temperature, even some new HNO3. In an imperfectly sealed canister the dT that produces "breathing" and its related vacuum could draw ambient air (with its few molecules of water) into the A.Nitrate tablet during cool-down. For the non-hermetic inflator, maybe there is a threshold number of pressure/vacuum cycles required to achieve a critical cumulative moisture in the tablet that, upon detonation, would promote inflator rate of gas expansion exceeding the design rate?