[Prev][Next][Index][Thread]

RE: Science Bashing?

 Hello,

   Now I didn't start this discussion.  If Chip wants he can toss
   this.  But perhaps I have some input of substance.

   I am not sure if either Mr. Hull or Mf. Leyh has it right.  There
   is some truth I see in both their statements.

   But certainly the statement of Mr. Hull is a little exaggerated to
   my way of thinking and my experience with physicists.  If it
   DOESN'T work on paper, it WON'T work in the real world.  Things
   have to work on paper first.  I think he was just talking about
   Tesla Coil building and operating, but his statements might be
   interpretted in the broader sense.


>Richard Hull wrote:
>
>> So much of theory is based on a supply of matched impedance or relative
>> stiffness and stability.  It is based pretty much on steady state (CW)
>> conditions. (not always) and relies often on neat sine wave input.

   OK, as far as it goes.



>> Violate one of these parameters and any good scientist will walk away and
>> you'll have to actually do the measurments, real-time!

   Now this is just too much of an exaggeration for us to swallow! 
   Transitory conditions in RF systems are *very* interesting, and I
   would not make that blanket statement that "any good scientist"
   will walk away.  Mr. Hull, while he does respect theory and he
   knows an awfully lot, thinks perhaps too little of theorists and
   too much of experimentalists.  (Would that be a fair assessment?) 
   Some of the most interesting things go on inside our cyclotron
   during transients!  Ever heard of multipactoring?


  Mr. Leyh wrote:

>Most 'theories' that depend on such a rigidly controlled set of initial 
>conditions are usually only 'rules of thumb' found in engineering handbooks 
>and technical manuals. 

   From one angle I understand and agree with this.  The circuitry
   that proves or demonstrates the theory is just too simple to be 
   found in the real world.  However, from another angle I think it is
   missing the point - the point is that to demonstrate a pure effect
   you have to keep out the contamination.  As Mr. Leyh hinted, this
   is important for the thought processes.  Why? Because the purity of
   thought is needed much as if you were to change two things in an
   experiment, then you really don't know which one caused the new
   effect, since there is no one-to-one purity.  Therefore, real-
   world electronic devices have been given 'equivalent circuits'
   which are collections of *pure* components, as complicated or as
   simple as necessary, to describe the device's operation in the time
   domain in question.  Then these equivalent circuits can be feed to
   a good circuit-simulation program.  Such simulations can be very
   accurate.

   The space program is a good example of how simulations have to be
   accurate.  And they also have a real model on earth to test what
   they can that way.


> These quickie equations are intended only to provide
>a general insight to a problem set, not to explain the universe in general.
>More accurate calculations of physical constants simply requires more work.

   True, but not the word "only".  Quickie equations are a way to
   simplify the thought processes for purposes of planning.  Or are we
   possibly saying the same thing?  I agree that to simulate the real
   world, it requires immense calculation power.  To simulate it
   perfectly requires the real device itself.

>A _good_ scientist, when confronted with a discrepancy in his/her model of
>a physical system will collect data, make a bold conjecture as to how the
>model should be improved, and then devise an experiment to test this conjecture.


   Correct.  Maybe even a couple of conjectures and test them both.


>There are times when it's quicker to use the 'cut and try' approach, but like
>any free lunch we should be constantly wary of the hidden costs.
>
>-GL


   In other words, if a person would just stop and really think about
   something, and if he has the proper background and schooling, then
   *maybe* he can figure out the answer in his head quicker than making
   up a real-world model.  Yes?  No?

   Well, maybe.  Or, maybe not :

   For instance, the analog (real-world) test-stand or model has its
   advantages.  Our big cyclotron here is a very complex RF cavity
   consisting of about 80 resonator segments grouped into 8 sections
   of 10.  That cavity is inside of another complex RF cavity (the
   cyclotron vacuum vessel).  Ideally we wish to keep the coupling
   between the two cavities as close to zero as possible so that 
   stuff in the outer cavity doesn't affect the tuning of the inner
   cavity, and the tremendous power inside the inner cavity does not
   leak out and melt things in the outer cavity just like cheese in a
   microwave oven.  With the tremendous RF power, we can melt lead,
   aluminum and copper like that.  I've done it.  The equations
   necessary to understand the screwball resonances were in the old
   days just too much to program correctly, and probably would take a
   long time to run.  (In fact that has been the bane of cyclotron
   design all along: insufficiently accurate computer codes.)  So, we
   built a 1/10-size scale model (a decision by Physicists and
   engineers) with a whole load of real, little, functioning RF
   probes, and externally adjustable resonator segments.   And it
   clearly revealed some coupling modes we didn't understand and
   allowed us to adjust things on the model to develop a tuning
   technique which would work on the big machine to control and
   optimize the resonant frequency of the inner cavity while at the
   same time minimizing the RF leakage to the outer cavity (our
   forward power meter is in megawatts CW).

   So modelling indeed has its place, as does pure theorizing.  The
   real trick is to know how and when to make the best compromise. 
   The theories tell one how to extract data from his models, how to
   vary the parameters to extract the maximum information in the
   clearest ways.  And good theorists don't run away from a problem.
   They might spout a lot of hot air, but not from running away,
   at least not usually.

   All the best,
 
 Fred W. Bach ,    Operations Group        | Internet: music-at-triumf.ca
 TRIUMF (TRI-University Meson Facility)    | Voice:  604-222-1047 loc 6327/7333
 4004 WESBROOK MALL, UBC CAMPUS            | FAX:    604-222-1074
 University of British Columbia, Vancouver, B.C., CANADA   V6T 2A3
 "Accuracy is important. Details can mean the difference between life & death."
 These are my opinions, which should ONLY make you read, think, and question.
 They do NOT necessarily reflect the views of my employer or fellow workers.