What's on my mind.

01 February 2008

Smashing the Lab

I've now watched two episodes of Discovery's new show "Smash Lab"*, the premiere and last week's (i think). The premise of the show, as far as I can tell, is these four people get together and try to invent something (or adapt an existing technology) to solve a problem, usually with much destruction of cars, house, etc. As Honu-Girl's husband described it "they're trying to be "Mythbusters" without any of the actual science." I love "Mythbusters" (partly because of the science, ok mostly because of it) and think "Smash Lab" could be good but Jamie and Adam may have to be come overhaul it. Allow me to explain. (It's my blog after all, how ya' gonna stop me?)

In the first episode they are trying to build a better median divider based on the arrestor beds at the end of airport runways. (They had a really nice video of a plane driving over/in one.) Some of the problems they encountered were due, no doubt, to budget and time constraints. And the nature of the material means that there isn't much one can do with scaled-down tests.
But I still have problems with their methods. First of all they decide to split up and try two entirely different methods. One group tests adding a 3-foot thick layer of the concrete foam in front of a standard barrier; the other decides that building an EMAS-type bed for the median would be the way to go. This divided their efforts and it felt to me like the producers, at least, were trying to make it seem like a competition.

The first serious problem I had was with their baseline crash. They all start out by doing a test of what happens when a car crosses the median at 60 mph and strikes on coming traffic at about a 30 degree angle (I may be misremembering the angle). Obviously it would be a huge logistical nightmare to have the oncoming traffic moving so those 6 cars are parked and only the median-crossing car is moving. They use accelerometers to measure the force of the impact. My problem at this point is that there is a huge difference in the energies involved in a wreck between one party moving and both parties moving. In their test the relative speed of cars is 60 mph; assuming in real life that the cars would be moving on the other side of the highway, the actual relative velocity would be almost 120 mph. I understand it isn't realistic to do a test under those conditions but I think they should have explained that a real crash would be different because those other cars are moving. They also spent more time than necessary talking about how the stationary cars had been moved and banged up in the test crash; more than necessary , IMHO, because again that is determined by the momentum of the vehicles as they enter the crash and their crash wasn't completely realistic. A nod at the differences between their crash and reality would have satisfied me.

My second big problem with the episode is that the first group testing the augmented Jersey barriers didn't run a test with the car just hitting the barrier. They went from no barrier to one with extra concrete in front of it, with no way of knowing what effect the barrier its self had on the situation. It was somewhat amusing that they succeeded in basically making a ramp for the car to jump the barrier, though. The other group's method was basically sound, I think, and with more time and/or money to try out more consistency of concrete they may have built a successful arrestor bed. They did, however, fail to think about the problems of pouring that much concrete, keeping it contained, and keeping the consistency uniform. I can forgive that, not many people have experience pouring a 60' x 30' x 2' bed of concrete foam.

The second episode I watched was them attempting to make an airbag for the front of a train to help reduce the force of impact of the engine with a car at 25 mph (That speed chosen because, apparently, at collisions at 25 mph and above fatalities become much more likely.) After a test crash (car parked on track) two go off in search of appropriate material for the airbags, two work out how to move the car off the track. The search for airbag material, how to fill it quicky using compressed gas (they didn't want to use the chemical reaction used in car airbags?) My main problem with the episode was the lack of true scale models. The two testing different ways to move the car to the side (a wedge, flipper, and pivot) tested them using a foot powered cart on a track and a club chair. A true scale test might have helped in final design and would have alerted them to the problem of how to fold the airbags up out of the way and hold them there. It might also have shown them that the pressure relief valves weren't big enough before the afternoon before the full-scale test. I understand that once they were at the site to set up they didn't have time to go back to the lab and think about these issues. I understand that sometimes you don't have the leeway to reschedule things (like the use of a track spur, train, stunt driver for the train...). The fact that they had to jerry-rig the valves and worked into the night on holding the airbags in place is a sign of poor planning.

Finally after the full-scale test they decide it was a failure and move on. One of the main reasons they decide it was a failure was because the car ended up flipping. It appears to flip because it hit a small embankment not directly because of anything wrong with the airbag system. There was an issue with the pressure relief valves blowing too late to be most effective, but that is hardly surprising considering the work around they came up with.

After all that rambling, the point is: I'm disappointed. It could be good. But the idea that one can successfully invent something using found materials for preliminary tests, no scale testing, and only one attempt at a solution is ridiculous. It might work if one is inventing a new cheese slicer, but one will never solve big problems that way. The show isn't a good example of how something is invented and it could be. But that might not have as many explosions.

*no link to "Smash Lab" because the site is blank. Apparently they haven't invented it yet, or maybe it was a "failure" because they screwed up and then couldn't go back and edit the code.

No comments: