[Sneap] Charging belts for AN2500

[Klaus Bahner]

Ian,



 
 
As far as I can tell, this means that when the last existing belt becomes
unusable, these small machines will all become useless, in spite of the fact
that there are still quite a number in operation after 30, 40 or perhaps
even 50 years of operation. If a small AN2500 or similar can still provide
the beams necessary for the research that is 
 
At least it is my understanding that it is not possible to convert the small
machines to a pelletron, due to space constraints and weakness of their
column structure.
 
What has so far been tried by those who no longer have a spare belt, or are
anticipating soon being in that position? What did the 'old hands' try all
those years ago - and what did they find that worked and what did they find
that didn't work? 
 
There were many experiments with "replacement" belts. In the past, these
experiments were supposedly driven by the wish to find a cheaper alternative
to the original HVEC belts. As far as I know these experiments were mostly
done on the larger machines, because lifetime of the belts on those is much
lower (by a factor 3-4) compared to the small machines. You may look up old
SNEAP proceedings and will find belts from companies like Amaraal, CIGO,
Wennerlunds and perhaps a few more. Results of these experiemnts were mixed,
some had succes, some saw failures - in my very personal opinion because it
is a time consuming and demanding task to experiment with belts. Most
research institutions have other things to do, than starting a charging belt
engineering enterprise - hence most of those experiments/solutions were
abandoned or only adopted by very few groups.
 
What really are the options concerning sourcing new belts - even if the cost
is quite high? Is it possible to imagine some kind of joint or cooperative
effort to search for and test various options? 
 
In my opinion, finding of any replacement for belts must be based on a
commercial, worldwide available solution. Any custom order fabricated item
has the major drawback, that it will be very expensive and only available
for the institution ordering the custom made belt. Even if you put all
existing belt driven accelerators together, the total required amount of
belt material is ridiculous low, to be of any interest for a belting
company. In my opinion, find something you can get off the shelf, otherwise
you are just asking for trouble.
 
In our case I used conveyor belts from a company called Forbo Siegling.
After many years of experiments in our EN (and *way too many* belt changes
:-(  ), I found a solution using one of their stock items running to our
full satisfaction. We are running now on the third finally chosen belts from
them and I dare say that I'm confident that the belt problem is solved in
our case. Still, this doesn't nrcessarily mean that this is an universal
solution. Running well in our case doesn't mean that it will work in any
other accelerators, too. This verification phase requires many more adopters
of this solution.
Today, there are a few, who use (or tried to use) these belts in other
machines, if my counting is correct two KNs, an AN, two vertical machines,
and another EN. I don't feel qualified to speak for those groups, so I won't
give any assessment on successes or failures here, but maybe those who tried
the Siegling belt may drop in and give their feedback. 
If you want to start to experiement with these Siegling belts, contact me
directly and I'll send you more details.
 


Over the years I have heard a number of stories - perhaps rumours myths and
legends - about belts : that performance was very batch dependant; that the
Passaic belts, in their brief existence, were rather mechanically fragile
and prone to tearing; that no-one really knows what specific characteristics
make a good belt - it is a 'trial and error' process to find one; that the
machines for making jointless canvas-and-rubber conveyor belts have now all
worn out to the point that they are not worth repairing just for making a
few belts for some old accelerators... 
 
Well, well, I don't know whether I just adding to the number of existing
myths, but I try to answer according to what I learned over the years
working with belts - but keep in mind, these are just my two cents. Feel
free to correct me.
 
According to old SNEAP proceedings performance was batch dependent.
Supposedly because the vulcanizing process was performed in small sections,
introducing the risk for making non-uniform belts. IIRC, Belt terminal
ripple spectrum was once analyzed by two guys from Utrecht, The
Neteherlands, I believe (Don't have the reference at hand right now), and
its spectral composition was both correlated to the weaving pattern of the
carcasse and the number of vulcanization segments.
Furthermore, as far as I know, the making of the rubber and the vulcanizing
process were not automated, which perhaps made it difficult to deliver a
perfect uniform quality distribution. Still, most of the belts worked well,
and actually I don't belive that problems with them were more pronounced
than with any other low volume technical product. 
What is a good belt? That is in fact rather easy to answer: Find a belt
which fulfills the mechanical requirements, i.e., proper dimensions, allows
running at high speed (without oscillations or deformations), excellent
tracking, resistant to wear caused by the charging system and has a long
lifetime in a somewhat hostile environment (dry atmosphere, possible crack
products of the insulating gas, high electrical field). The electrical
requirements are, finding a material that is highly resistive, causes low
self-charging, can withstand the local fields and has roughly the same
capacitance, if you don't want to change your charging system drastically.
It's getting difficult due to the fact that those belt parameters, both the
mechanical and electrical ones are ill defined and are usually not easily
accessible by measurements. Once mounted in an accelerator the possible
number of measurements you can perform in order to explain a belt's behavior
are very limited and always only indirect. 
Besides that, I believe that any material fulfilling the above requirements
can be used as charging belts - especially there is nothing special about
the canvas-neoprene combination. Rubber has the advantage that it is still
the best suited material, when it comes to withstand abrasion by the screens
and a canvas carcasse was in my opinion chosen because this was a well
established technology for endless belts (used as power transmission belts)
in the first half of the last century, read as: it was rather easy to source
the endless carcasse in the various required dimensions. Having an endless
belt is a good starting point in order to fulfill the mechanical
requirements.
Most power transmission belts (with the width typically found in
accelerators) where either replaced by gears or more powerful synthethic
fiber products. Hence you won't find many vendors still delivering them. You
still can get endless belts (on non canvas carcasse) with a rubber coating,
but not with the required dimensions, especially not belts with a thickness
of only a few millimeters. They are for example still used in mining
operation, but of course they are much thicker there, than what could be
used in an accelerator. In conclusion, if you need a thin, affordable and
off the shelf belt, that belt will not be endless.   
The next major difficulty in finding a suitable belt material is the
requirement that it must be non-antistatic, i.e. highly resistive. For most
industrial applications of conveyor belts this is an intolerable
requirement, because static electricity which is generated, when the belt is
moving, causes a safety hazard. Hence the majority of all modern belt
materials is conducting and hence by principle useless in an accelerator.
Only for few niche applications (e.g. electrostatic coating) non-antistatic
belt material is produced and only a few vendors have such belts in their
product portfolio. 
 
 Thanks for sharing your thoughts and experiences 
 
Well, the above are just my (possibly wrong) thoughts,  but I hope it is at
least somewhat useful.
 
Kind regards,
Klaus Bahner
--------------------------------------------
AMS 14C Dating Laboratory
University of Aarhus
Ny Munkegade, Byg. 520
DK - 8000 Aarhus C
Phone: +45 8942 3736
Fax:   +45 8612 0740
-------------------------------------------- 
 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://daytona.tunl.duke.edu/pipermail/sneap/attachments/20070614/e2ea2c76/attachment.htm