Chapter 6

[quixote]

I think the relevant comparison here is not how fast the density of the plasma dropped off, but how slow it dropped off compared to the current being used to power the plasma source.

[mvanwink5]

I suspect several of you are madly counting CF flange bolts and looking up sizes in the catalogs. That will keep you busy for a while, but I notice the camera angles don't show a lot.

Above, 44 but I'm poor at the catalog search.

I took another look and compared the pictures showing the large vacuum flanges on page 66 and 67 and I recounted bolts. As I count, there are 48 bolts per large flange, the same as in the EMC2 website shown in the sketch labeled WB-8. That would imply that the EMC2 drawing is indeed WB-8.
Best regards

[mvanwink5]

In looking for flanges I found a manufacturer of e-guns. Funny, but the sizing terminology reminded me of the Dec 2011 .gov report when the e-gun size was mentioned. It seems that sizing of e-guns is given in terms of "heating," so "heating" was not referring to what EMC2 was trying to do with the larger e-guns, it just meant they were buying larger e-guns.

http://www.hositrad.com/vacuum-products/e-guns/

For purchasing the type of connecting flange, the name is a "cluster flange" here is the link to the catalog showing the flange design, see page 30

http://www.hositrad.com/docs/upload/119.pdf

Still haven't got anywhere with guessing size though....

Best regards

[John Gallagher]

The view ports on the chamber have a standard conflat arangement and are 6" OD. Scale from there and you get a 24" cluster flange. The smaller 6 bolt flanges are 2.75" OD

[John Gallagher]

A close match would be a wire seal flange with an OD of 27.125" . Kurt Lesker and Norcal list them as having 40 bolt holes.

[mvanwink5]

Thanks John for the help. Back to 30 cm coils then for WB-8 which is smaller than we were thinking. Back to WB-7 size coils, and would explain why EMC2 would use a WB-8 labeled drawing for the WB-7.1 device on their web site.
Best regards

[MSimon]

A close match would be a wire seal flange with an OD of 27.125" . Kurt Lesker and Norcal list them as having 40 bolt holes.

Could be a custom job.

[D Tibbets]

Actually, the shutting off of the plasma injection was kind of the point of that figure, as I understand it. When operating at as close to steady state as they operate the device, they shut off the 'plasma source' (they don't say what it is), and monitored the loss of density of the plasma.

90% of the plasma was gone within the 0.2 ms of stopping plasma injection.

My question is, does that tell us anything about confinement or other metrics of the device if the dwell time in the machine is less than 0.2 ms for a typical particle? Or am I misinterpreting that figure?

Determining the significance is possibe, at least in a rough fasion.A simple and perhaps acurate analysis follows.
Assumptions:
Assumed drive voltage of ~ 10,000 volts gives helium ions (helium often used for such studies) a speed of about 100,000 M/s or 10,000,000 cm/s.
Magrid diameter = 60 cm.
10,000,000 cm in one second means it would take ~ 60 cm/ 10,000,000 cm/s = 0.000006 s to transit the machine. This is 6 micro seconds.
Now the fuzzy parts. Thermalization issues and scattering issues may change this number for the average ion. Assume that the number may be 2-3 times as high. Also consider that not all of the ions will be on the edge of the Wiffleball and flying towards the other side. Many will be near the center, so using radius instead of diameter may be more representative. This will lower the non confined transit time. Also the loss rate may not be linier. I've thus estimated a very uncertain unconfined transit (escape) time of 5 micro seconds. Most of the ions would leave in this time.
If the measured time till most of the ions had left was ~ 200 microseconds., then ~ 40 transits would have occured on average.
The EMC2 pattent application mentions that no confinement = one pass befor loss. Mirror confinement was estimated to be ~ 5-8 passes or confinement time of ~ 30 microseconds.
Cusp confinement (Polywell confinement without Wifflebal contributions) was estimated at ~ 60 passes or ~ 300 microseconds under these conditions.
Note that Wiffleball confinement increases this by a factor of perhaps 50, so the confinement would be ~ 15000 microseconds. This is not much different from some numbers I have seen.

My conclusion based on these uncertain but conservative assumptions is that a 200 microsecond confinement time is consistant with predictions for cusp confinement times for the Polywell.
Actually, using the ion speed may have been innapropriate. Using the electron speed may be more pertainate. For that the transit times would have been closer to ~ 0.1 microsecond (~1 /60 times or more rapid). The 200 micro seconds would thus represent ~ 2000 electron passes. This is important because the ion confinement is not based on the magnetic confinement as is the electrons. The plasma would have to be near neutral, so electrons would have to be present, though the source apparently was not given.
The 2,000 electron passes confinement time (0.2 ms) under cusp confinement conditions would seem to indicate much better performance than predictions. That, or if the predictions are based on WB6 or 7 conditions, then the size and /or the magnetic scaling under WB8 conditions are showing significant improvements.

Dan Tibbets

[D Tibbets]

Thanks John for the help. Back to 30 cm coils then for WB-8 which is smaller than we were thinking. Back to WB-7 size coils, and would explain why EMC2 would use a WB-8 labeled drawing for the WB-7.1 device on their web site.
Best regards

I don't follow. The upper limit on the magrid size is limited by the largest opening into the Vacuum chamber. This has been hashed out to be ~ 60 cm. This of course assumes that each separate magrid is placed on it's own support. The assembled magrid as in WB6 is probably a thing of the past.

Supporting this is some drawings, and also stated goals. WB 6 magrids were stuffed with wire and a little electrical insulation. There was no room for cooling plumbing. There can be some argument about The thickness of the coils. WB4 coils were about 25% thick compared to machine diameter, WB6 was ~17%. And WB4 had water cooling. Still I suspect the size is larger than 30 cm. There is more room for everything.
The only sure thing is that the maximum diameter cannot exceed ~60 cm unless the vacuum chamber was built around the assembled Polywell.

Dan Tibbets

[mvanwink5]

This has been hashed out to be ~ 60 cm

Thanks Dan, do you remember where this was done?

[KitemanSA]

Search on me as author and "conflat" (thanks, Tom L), and you should find the post pretty quick.

[Robthebob]

I'm reading through the paper, and to my surprise, its about plasma Interferometry. I just so happen to have done plasma interferometry for undergrad research at Auburn. Someone beat me to the punch, sigh...

I need to go to New Mexico for grad school it seems.

[mvanwink5]

Thanks KitemanSA, that was very helpful, somehow I missed those posts before. On the page 66 picture in chapter 6, the flange bolt counts match the WB-8.jpg labeled picture on the EMC2 web site for the flanges I could count bolts on including the 48 bolt major flange, so the dimensions should apply.
Best regards

PS for those interested here is a link
viewtopic.php?p=38477&highlight=60cm#38477

[erblo]

After looking at the pictures of the chamber (6-1 and 6-2) I have a few questions for those with some experience with vacuum systems involving high voltage and/or electron/ion guns (or anyone who has an opinion ) :

Does the white rod+box labeled "WARNING HIGH VOLTAGE" seen on the top left of the big center flange (on both opposing sides and possibly on the one to the right in 6-2) imply independent coil voltage supply for each coil, i.e. no nubs? I'm assuming that this is a high voltage feedtrough, could it be something else?

The other thing present on both sides, although in different locations and configurations, is the assembly to the right of the center in 6-2 (middle of the picture). Any ideas on what this is - gas puffer, plasma source...?

How visible should the coil current feedtroughs be? Does the absence of big cables connected to the side seen in 6-2 tell us anything (or could the coils just as easily be connected inside the chamber)?

How much individual control of the coil voltage/current would be feasible/desirable? This is based on the sentences "This particular shot used all 6 coils but no electron gun. /.../ For this shot the coils were not charged." - which most likely is just the way he wrote and doesn't mean anything
Thanx!

[Northstar]

Actually, the shutting off of the plasma injection was kind of the point of that figure, as I understand it. When operating at as close to steady state as they operate the device, they shut off the 'plasma source' (they don't say what it is), and monitored the loss of density of the plasma.

90% of the plasma was gone within the 0.2 ms of stopping plasma injection.

My question is, does that tell us anything about confinement or other metrics of the device if the dwell time in the machine is less than 0.2 ms for a typical particle? Or am I misinterpreting that figure?

Determining the significance is possibe, at least in a rough fasion.A simple and perhaps acurate analysis follows.
Assumptions:
Assumed drive voltage of ~ 10,000 volts gives helium ions (helium often used for such studies) a speed of about 100,000 M/s or 10,000,000 cm/s.
Magrid diameter = 60 cm.
10,000,000 cm in one second means it would take ~ 60 cm/ 10,000,000 cm/s = 0.000006 s to transit the machine. This is 6 micro seconds.
Now the fuzzy parts. Thermalization issues and scattering issues may change this number for the average ion. Assume that the number may be 2-3 times as high. Also consider that not all of the ions will be on the edge of the Wiffleball and flying towards the other side. Many will be near the center, so using radius instead of diameter may be more representative. This will lower the non confined transit time. Also the loss rate may not be linier. I've thus estimated a very uncertain unconfined transit (escape) time of 5 micro seconds. Most of the ions would leave in this time.
If the measured time till most of the ions had left was ~ 200 microseconds., then ~ 40 transits would have occured on average.
The EMC2 pattent application mentions that no confinement = one pass befor loss. Mirror confinement was estimated to be ~ 5-8 passes or confinement time of ~ 30 microseconds.
Cusp confinement (Polywell confinement without Wifflebal contributions) was estimated at ~ 60 passes or ~ 300 microseconds under these conditions.
Note that Wiffleball confinement increases this by a factor of perhaps 50, so the confinement would be ~ 15000 microseconds. This is not much different from some numbers I have seen.

My conclusion based on these uncertain but conservative assumptions is that a 200 microsecond confinement time is consistant with predictions for cusp confinement times for the Polywell.
Actually, using the ion speed may have been innapropriate. Using the electron speed may be more pertainate. For that the transit times would have been closer to ~ 0.1 microsecond (~1 /60 times or more rapid). The 200 micro seconds would thus represent ~ 2000 electron passes. This is important because the ion confinement is not based on the magnetic confinement as is the electrons. The plasma would have to be near neutral, so electrons would have to be present, though the source apparently was not given.
The 2,000 electron passes confinement time (0.2 ms) under cusp confinement conditions would seem to indicate much better performance than predictions. That, or if the predictions are based on WB6 or 7 conditions, then the size and /or the magnetic scaling under WB8 conditions are showing significant improvements.

Dan Tibbets

Dan, thanks for this analysis. While there is a lot that is not made clear in the background to this figure, your analysis is consistent with the 'very good confinement' report from about the same time.

Clearly this thesis take care to not publish much that tells outsiders much of anything as to how WB-8 is working. The inclusion of this figure almost stands out to me as someone 'throwing us a bone' of some information if we are able to figure it out.

I'm sure that there is more in there to dig out of the other figures and data, if we are able to work out what it is.