Revisiting an old topic - Wall loading

It may be difficult to separate "theory" from "application," but let''s see if this helps facilitate the discussion.
Post Reply
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Revisiting an old topic - Wall loading

Post by Richard Hull »

Once again, as the HEAS event approaches I am warming up the idle fusor IV. The usual three day plus warm up began on Monday. I hope to continue to run it daily until the Friday pre-event gathering. I will not push it hard, lest something go "high order" before the event.

I posted about this last year I believe and checked against my data logged values to compare results with this year.

On Monday, day one, I evacuated the fusor in about 30 minutes to unload all water that usually accumulates in the system and oils during months of non-activity. I started the fusor with a base level of 4X10-4 torr. (.4 micron) The most Deuterium pressure I could get to function was no more than 5.2 microns and this was after 2 more hours of running and 8, 10 minute runs over that period. I took data for all runs over the last 5 days. I will give only the data for the first run of the day and the best, (last), run of the day.

day 1 First data run 4.4 microns D2, 38.5kv,10.5ma, 3,000n/s....Last run 5.2 microns D2, 40kv, 10ma,24,508 n/s
day 2 First data run 5.5 microns D2, 39kv, 6.8ma, 62,500 n/s....Last run 7.6 microns D2, 40kv, 10.5ma, 195,200 n/s
day 3 First data run 7.5 microns D2, 37.7kv, 9.1ma, 223,400n/s...Last run 8.1 microns, 39.1kv, 10ma, 326,600n/s
day 4 First data run 9.5 microns D2, 38.2kv, 9.2ma, 410,412n/s...Last run 11.3 microns, 40.5kv,10ma, 746,900n/s
day 5 First data run 12.3 microns, 41.8kv, 9.1ma, 928,641n/s....last run 13.3 microns, 42kv, 9.3 ma, 1,021,000 n/s

A check of last year's, (2013), first three warm-up days. Best runs of each day, only

day 1 Best run took 5 hours! 6.5 microns, 35kv, 10.7ma, 33,600 n/s
day 2 Best run 6.6 microns, 39kv, 7.2ma, 111,600 n/s
day 3 Best run 9.1 microns, 38kv,11.5ma, 350,000 n/s

In both years, the second and third day's runs were completed from evacuation to final run in less than 2 hours.

I was amazed to see similar consistencies year to year for the past 5 years that I have kept such records. Always a rotten first day and many hours spent nursing a forlorn hope for that day. The second day is always better and things move right along. The third day and all subsequent days only an hour or more, at most, yielded fabulous, easy runs.

I still think on that first day the dormant fusor has soaked up a lot of moisture and this must be dispatched, regardless of similar starting pressures day-to-day. On that first day, the fusor's shell starts to get loaded with fast neutral Deuterium gas atoms. Subsequent runs, on later days, start not only with the same or very similar D2 pressures, but with a loaded wall which pours out its bounty of stored D2 atoms at the ideal shell distance as it is bombarded during operation and many of these perfectly placed "wall squeezings" are readily ionized at the ideal "acceleratory spot" in the fusor volume by the flood of electrons heading to the shell.

A real tell on this is that in the early runs, as I up the voltage and current, the D2 pressure actually drops! Sometimes this drop can be .5 microns over a 10 minute data run. By day three, the same raising of voltage and current always raises the pressure by no more than .2 microns, but it rises, nonetheless. Wall D2 being bombared out of the shell?

Stainless steel is not a great hydrogen store, but all metals are eager to absorb or load hydrogen under the right conditions. This is a well known fact. Nickel is a great hydrogen store, but there is not much in 304 SS, but perhaps enough to see in experiments such as this.

Now, have any of you old boys, who still run your fusors, seen this effect in your fusor? Note: It must be idle for a long period and suddenly run on successive days with good data taken. There are very few still out there reading this that can respond with data, I am sure.

I gotta line my chamber with either Ti, Ni, or Pd foil!

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Revisiting an old topic - Wall loading

Post by Richard Hull »

I have updated the above post to include day 5. Fabulous results. The mega mark! Fusor IV is ready for HEAS.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 1:27 am
Real name:

Re: Revisiting an old topic - Wall loading

Post by prestonbarrows »

Can't say personally for fusors but beams onto solid targets have similar behavior. It varies, but loading up the target takes on the order of hours of run time.

On the other hand, conditioning out water and such takes on the same order of time. It could be that even at the same supply current, you have less D current since contaminates are in the mix adding overhead.

Also, It is very possible you are also loading up your grid. However if your grid is getting over 200-300 C, any hydrogen will quickly diffuse out and there will not by much of a net effect.

If you want to narrow down what is actually happening, try venting to atmosphere after some baseline conditioning. This will not greatly affect the D loading but will reintroduce water into the system. A few cycles of this will show if your yield is scaling with plasma conditioning time on each cycle or overall surface loading time across multiple cycles.

If you are actually target loading, titanium will give many times the yield of stainless or pretty much any other material. A gate valve covering a hunk of Ti giving a repeatable change in yield would be an interesting confirmation of wall loading.
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Revisiting an old topic - Wall loading

Post by John Futter »

Preston
I do not know where you get the hours from!

I regularly put lots of atoms into various targets.
It is easy to work out when enough is enough
Target current gives atoms arriving per second ( you have to suppress secondary electrons to get this right)

but usually after 10^ 19 atoms per sq /cm you will go past 50% implanted and the target will delaminate with gas ions.

so lets have a look at a typical fusor
40 kv 10 mA with an ion source beam diameter 2cm on target


so 1^2 x pi = 3.1412
current = 1.6^17 atoms /s

which = 0.5 ^ 17 atoms / s /cm
so to go over 10^19 atoms 200 seconds =3.33 minutes


we use charge integrators to measure this directly
but on new EQ without we use mA/minutes and the known implantation area.
new EQ we usually have no idea as to the implantation area so we put a whole 8" wafer in and let it run for 15 minutes at around 10-20mA ion current.
Taking out the wafer you can easily see the ion damage as the surface is delaminating, even a blind person can feel the damage area.

In the case of fusors and because of hydrogen and its isotopes diffuse through the implanted surface to escape the damage is hard to see. The same applies to helium.

Argon is much better and will give easy to see damge trails (and its cheap)
prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 1:27 am
Real name:

Re: Revisiting an old topic - Wall loading

Post by prestonbarrows »

John Futter wrote: I do not know where you get the hours from!
From personal experience operating deuteron beams into titanium targets as well as the published literature.

Since you mention delamination and wafers, I assume you are thinking in the context of the semiconductor world where exposure times are indeed quite short.
John Futter wrote:In the case of fusors and because of hydrogen and its isotopes diffuse through the implanted surface
Absolutely correct, that is what leads to the relatively long loading times.

When you dig into it, there is all sorts of materials science going on with different chemically bound phases of titanium hydride and free hydrogen states in dynamic equilibrium. You can approach the range of 200% D to Ti ratio. This is all affected by temperature and the grain structure of the Ti and other voodoo. I won't pretend these details are in my area of expertise.

For the original context of fusion from a D beam into a Ti target at 10's-100's of kV, loading will generally be on the order of hours characterized by yield which steadily rises then plateaus somewhere many times it's initial value. Resting the system over night leads to a fractional loss in yield the next day which quickly recovers (again, explained by diffusive losses of free D in the lattice structure). More current generally leads to higher diffusion rates and the time to plateau stays in the same ballpark for all beam systems I have seen.

Ti is a bit of a special case because of its chemistry, but most things have roughly the same behavior just with scaled down yield.

Here is a quick example from publicly available literature. Relevant parts around page 100.

https://escholarship.org/uc/item/35w719tr.pdf

Image

Or another example behind a pay wall
A comparative study between pure titanium and titanium deuteride targets used for neutron generator wrote:The neutron yield as a function of bombarding time is presented in
Fig. 2. It can be seen that the neutron yield of pure titanium target
exhibits a gradual increase with the evolution of time. This is easy
to comprehend as the implanted deuterium atoms are accumulating
within in the target and chances for the D(d, n)3He reaction are
therefore multiplied.
Image
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Revisiting an old topic - Wall loading

Post by John Futter »

Thank you Preston

I am quite aware of these graphs and that they were done in the ones shown outside of where the amateur efforts lie.

indeed as the energy go up the implantation depth goes up and the consequent straggle of of ion depth rises exponentially requiring longer implant times.

at typical fusor voltages ie 20 -60kV implant depth will be below 100nm with no more than 20nm of straggle so it is easy to get past50atomic % loading very quickly at 120kV depth is now 300nm with 100nm of straggle and yes will take hours to get above 50 atomic percent due to the volume presented by the straggle..

I did model this for Steven Sesselman some years ago for his fusor I will put up the plots when I find them
prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 1:27 am
Real name:

Re: Revisiting an old topic - Wall loading

Post by prestonbarrows »

I would be interested in seeing. Most of the systems I have had direct contact with have been in the 50-300 kV range.

Did you model these in SRIM or something else?

Of course, in the context of a fusor, the picture is quite different. The fast-neutral equivalent currents into the walls will be spread across some semi-thermalized energy distribution and corresponding spread of implantation depths. The 'target' surface area is also relatively very large and at a mix of angles of incidence to the 'beam'. Under these conditions, I could only speculate what the cumulative end result will be.
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Revisiting an old topic - Wall loading

Post by John Futter »

Preston
Yes I did do the target loadings in SCRIM for Steven as his system was beam on target
we also use dynamic TRIM which models the damage cascades much better and these agree with our RBS measurements as to depth and concentration especially when the ion energy goes above 40kV whereas SCRIM starts being inaccurate as the ion energy goes up.

As to most Fusors I think of them as multiple beam on target devices as evidenced by the lighting of the neutrals / d2 molecules by energetic ions in the star pattern formations.
It would be very easy to prove this by someone putting SS foil in contact with the shell and running, then Deuterium profiling the foil with RBS or RNRA to prove areas of increased concentration. As no one has done this yet it is still in the arena of armchair conjecture

work for the future
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: Revisiting an old topic - Wall loading

Post by Chris Bradley »

John Futter wrote:so lets have a look at a typical fusor
40 kv 10 mA with an ion source beam diameter 2cm on target


so 1^2 x pi = 3.1412
current = 1.6^17 atoms /s

which = 0.5 ^ 17 atoms / s /cm
so to go over 10^19 atoms 200 seconds =3.33 minutes
hmmm....

but a 10mA fusor has an electron current of 9.83mA and an ion current of 0.17mA. == 200 minutes?

Your 10mA beam currents are 10mA of ions, John. A regular fusor is all electron noise with a statistical anomaly of moving ions.
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: Revisiting an old topic - Wall loading

Post by Chris Bradley »

I think my opinion on wall loading in a fusor is well known now, but I'll just add it to the 'information point' here - if there were significant beam-beam or beam-background reactions I should have seen lots of neutrons with my epicyclotron. I didn't. I built the whole thing assuming it was the latter. It turned out to be neither. Mine was the 'null-hypothesis' test for the colliding beam theory of the fusor.

It is a bit odd perhaps that a 'non-fusor' experiment would help explain the fusor. What I did see was a huge loss of ions through charge exchange. DD fusors are almost entirely about fast neutral bombardment into the walls, IMHO, in which case wall loading, time to load, material type, maximal interstitial densities, and beam focus would be the significant factors in reaction rates.
User avatar
Garrett Young
Posts: 123
Joined: Fri May 26, 2006 9:51 pm
Real name: Garrett Young

Re: Revisiting an old topic - Wall loading

Post by Garrett Young »

Richard,

With each subsequent day during the five day period you mentioned, were you able to achieve a lower base pressure before back filling with Deuterium? Or were all days the same as the first with 4X10-4 torr (.4 micron)?
- Garrett
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Revisiting an old topic - Wall loading

Post by Richard Hull »

I have posted long detailed lists in the past of multiple day runs. The fusor needs conditioning in that the more it is used in frequent daily runs it just gets better with each use. Extraneous material is driven from the walls via bombardment of successive runs allowing a bit lower pumping to start. However, as the runs progress on any single day, the D2 pressure can be increased consistently over that run for better results. I do not believe I ever ran the system over more than 10 consectuive days and may have never reached what I felt might be a "wall loading" limit.

Operation of a fusor is something you get a "feel" for, viscerally, but only over time........A long time. Each fusor, I believe, has its own operational regime. I feel this is due to the amateur construction, assemblege, and a learned understanding of what makes each one perform.

I also believe that any still practicing "old boy" fusioneer could go to any fellow fusioneer's system and run it well after getting a feel for it.
Only a man who has built one from scratch and toughed-out through the learning period can take the wheel and begin to make one perform to its highest level. This is because the fusor, as we build it, is not elegant, but a bull-headed system and you have to learn how to herd and drive the bulls.

I, again, feel that wall loading is a major contributor in ever more successful, protracted, successive runs. Walls can only load slowly due to fast neutrals. I think the increase in performance over time is soley due wall loading and re-emission of deuterons at the ideal place in the electrostatic velocity space. (the walls/shell) One need not fiddle with ion guns with their hyper thin beams when every square cm in the inner spherical surface area is deuterium loaded and via electron bombardment, bouncing out deuterons capable of full acceleration.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Post Reply

Return to “Fusor and/or General Fusion Theory (& FAQs)”