Hacker News new | comments | show | ask | jobs | submit login
Analyzing Radioactive Glass from the Trinity Test (hscott.net)
76 points by blueintegral 1607 days ago | hide | past | web | 14 comments | favorite

The Trinity site is open twice a year to visitors. When I went 10 years ago there were still bits of the blueish-green glass mixed in with the grass and sandy soil. A small section was preserved and coverered with a low windowed box. The ground inside was entirely covered with the strange material.


I tried to do this tour about 10 years back - be aware that there are some quite u expected requirements for non US nationals wanting to do the tour. (or at least there were back around 2000/2001)

Still disappointed I didn't manage to organize that…

As an undergrad at Georgia Tech you should know that if one of my professors caught me writing some significant digit nonsense like 1.0272303 ± 1.031768 Bq/kg I would not have gotten a degree.

1.0272303 ± 1.031768 Bq/kg or rather 1.0 ± 1.0 Bq/kg means that your experimental error is so large that your result is compatible with both 0 and 2 Bq/Kg. What's the point of adding all those significant digits if you're telling us that your result is imprecise, or rather that your experiment can't effectively tell if there is any activity from Cs-60.

(edit: I wasn't an undergrad at GT)

That data came from the HPGe, which really does have that precision. I know it seems like the accuracy is bad since, like you said, it could be either 0 or 2 Bq/kg. However, a Bq/kg is REALLY tiny and that accuracy (which is proportional to the length of the experiment) is good enough for this analysis. So the precision is fixed based on the HPGe and the accuracy increases with the length of time the HPGe took measurements.

My point stands: just write 1.0±1.0. If 1 Bq/Kg is small then giving a precision of 0.1 nBq/Kg is just plain insane.

You're right, I'll correct it.

That's great! I see you're not a physics major, so I apologize if I came out as aggressive.

In general you want to keep 1 or 2 digits in the s.d. and align the precision of the mean accordingly.

I thought this exact thing. When I was at GT, this would have resulted in a grade of 50 +/- 50. Once you get to that error, you need to redesign the experiment.

Both of the videos at the end of the article are absolutely fascinating (as is the rest of the article, but I almost skipped the videos; I'm glad I didn't).

The first video is amazing. The structure to the explosion is so beautiful, but at the same time, I feel like I'm looking at the face of death.

I actually stopped watching git 10 seconds or so in, assuming the whole thing was just going to be Ken Burns-ed stills of photos I'd already seen. Thanks for making me go back and watch it all…

I'll also drop in a recommendation for the boom "100 Suns" for anybody who finds pics like that beautiful...

Wow, amazing

So I suppose what points this to being Trinity is the ratios, since the values are very off from the Trinity data in the tables (because of decay)

A Quote: "Kenneth Bainbridge, director of the Manhattan Project, was not amused with Fermi scaring all the guards."

Correction. Leslie R. Groves, Jr. was director of the Manhattan Project.

Thanks, fixed.

Applications are open for YC Winter 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact