Bob Greenyer comments that on the last 2 MFMP experiments, the resulting
sintered Ni + Li, Al mass slid out of the alumina tube with no apparent
sticking/sintering to the alumina.  Thus, a reactor that is safely
open-able after the reaction provides opportunity to sample the ash, and
with the correct plumbing, the gas may also be sampled.  Subsequently the
same reactor could be used again.

Since the plan to use the "easier Parkhomov design" with the Swagelok
termination uses substantially off-the-shelf components, and no glue
(reproducibility issue), the replication can be done many times with less
speculation of what changed between trials going forward.  OTOH, it is hard
to guarantee reproducibility if glued end seals are used.  Parkhomov is a
careful experimenter and he seems to have mastered the seals.

Alan Goldwater's tests of the seals have really moved this project forward.

On Sun, Feb 8, 2015 at 11:25 AM, Bob Higgins <[email protected]>
wrote:

> For some of the deviations, there was no good excuse.  For some, there is
> good reason.  MFMP has had difficulty replicating the Parkhomov seals, and
> does not yet have the right size alumina parts for proper Parkhomov
> replication.  And in the Parkhomov design, it is hard to tell if the seal
> failed or the device just didn't work.  The best data from the experiment
> just completed is that the sealing of the compression fitting with the
> aluminum ferrule was good.  This is a serious win, because it permits
> analytical plumbing to be attached to a Parknomov-like device.  This
> analytical plumbing will allow real time pressure measurement, post
> experiment gas sampling for analysis, gas venting to safely open the
> reactor, and an ability to open the reactor without cutting the tube (but
> it is not clear that it will be possible to sample the Ni materials without
> sectioning).  And, the plumbing will subtend no significant expansion
> volume.
>
> MFMP is getting ready to publish my proposed plumbing design for attaching
> to a Parkhomov-like reactor, probably on Facebook.
>

Reply via email to