Friday, November 21, 2014

Checkstop - Repeat Telecast

******************************************************************************************************************************






I then graduated to the exalted position of a CSIR Junior Research Fellow in the Nuclear Resonance Labs at AU, Vizagh in 1963.

And was attached as a helper to the senior-most scholar on the brink of submitting his thesis. It was an experimental work but he wanted to include some calculations to buttress it. And asked me to go the Library and fetch the celebrated tome on the Crystal Structures by Wyckoff.

Then he opened the book and asked me to take down the structure of his crystal with distances in Angstroms. The idea was to calculate the electric field gradient at the site of the Chlorine atom produced by as many near neighbors as I could include. He used the point charge model, as crude an approximation as treating a frog as a virus. But it was new and it would do for a starter.

And he gave me permanent access to one of the two primitive mechanical Facit Calculators in the lab:


To add, subtract, multiply or divide numbers, you have to crank the handle n number of times. He also gave me plenty of drawing sheets, pencils and erasers and asked me to get going. 

I felt like that Jabez Wilson in the Redheaded League.

Recall that the electric field gradient is a bloody tensor with a merry number of components. What does a senior research scholar care for the travails of a fresh entrant? 

It took ten days of 10-hour sittings for me to calculate the damn thing to its nearest neighbors and three months to the next nearest neighbors.

I then protested and gave up...this was not my idea of research in dream physics. And yet the chap was happy and included my labor in his thesis and thanked me for my sweat in his thesis as well as three of his papers.

And when I left my research position at Vizagh midway in 1965 and joined IIT KGP as a teacher, I wished to do a bit of my own field gradient calculations on the crystal KBF4 (Potassium Tetrafluoborate for you). Not with the museum Facit machine but with the newfangled IBM 1620. So I learned programming in Fortran II and tried it out first with the nearest neighbors.

The waiting time for a slot in the hallowed IBM machine was 3 months but the thing did it in 2 minutes flat after all the punching, pre-compilation and compilation rigmarole was through.

I was delighted with the printout of the components and became greedier and asked the machine to do it with the next nearest neighbors. The machine obliged but took half an hour. 

And then I became even more greedy and asked it to do it with the next to next nearest neighbors.

It went into coma for two hours at the end of which there were red red red lights all over with a screaming warning:


CHECKSTOP


I went home and examined the printouts and found that the numbers were oscillating...I mean they were not converging with increase in the number of neighbors included...no way...

Love's Labor Lost!!!

Keep it under your hat...


 The IBM 1620 Motto

Input Bullshit...Output Bullshit






...Posted by Ishani

*****************************************************************************************************************************

No comments:

Post a Comment