A nuanced approach to EMC2

[billh]

Tom Clarke,

Thank you for a calm and fair summary of the situation. I agree with you completely. I think the anger at Rick Nebel is completely unjustified. He hasn't over-promised and he hasn't lied. He just...hasn't said much. It always seemed likely that the only way we would ever learn anything about the results is by inference from whether the Navy keeps funding the next round of experiments. That was true before the FOI and it's still true. Nobody ever promised us fan-boys any more than that. I'm willing to wait. I may as well be, because getting mad or impatient isn't going to accomplish anything anyway.

Bill

[D Tibbets]

How or from where did you arrive at the ~500W?

By other posted analyses, WB-6 = 0.6mW.

WB8 has 8 x B field and 5/3 x size.

.0006 X 8^4 x (5/3)^3 = 11W.

Dang, how did you get 500W?

One perameter left out- the drive voltage. A rough estimate from graphs. Increasing the voltage from 10,000 volts to 100,000 volts will increase the crossection by ~ 30X. And, multiply that by any gains from an updated design, such as elimination of nubs. It would be a big change from 11 W to 500 w if it was due to the nubs alone.

Dan Tibbets

[Aero]

5/3 radius increase. Where did you get that and please don't say from the picture on EMC2's website.

Well if you didn't want the answer, why ask the question?

I have looked long and hard at that image, I've even used pixel calipers on it and I just don't see any way to get valid information. I know Tom said that the 30 hole ring seal was 14 inches (I thought) but that doesn't help when there is no view point to the drawing. Or maybe it does. How? Please spell it out for me.

[TallDave]

Re 500W -- I started with my old .002 watts from WB-6 (which may be optimistic) I got when I was discussing with Art on the PW wiki back before Talk-Polywell was around (I was trying to figure out if the PMTs could have been picking up fusion energy, an amusingly futile exercise in retrospect), in those desperate days before the WB-7 contract when it seemed this idea would never be funded. I then gave it a pretty generous kick: .002W*8^4*2^3*10 to get ~500W -- give or take an order of magnitude. It's hard to know what they'll be doing in terms of voltage and improving ion focus. I wouldn't be too surprised if they could get a kW out of it, if theory is sound and they really wanted to show off.

I don't think the nubs really make any difference to power. You could get the same fusion power either way, because current isn't the limiting factor, the B field is. It would just cost you more current to operate at that density, if losses are higher with nubs. (Obviously losses matter a lot more at, say, 10MW input power, esp. if you're hoping to get 100MW output).

But what can be done with a small machine that could not be done with a larger one?

Time and money, and as you say, political capital. As to the issues that can be resolved -- there are a lot of questions specific to p-B11, but electron loss scaling at .8T is probably the big question. Except for the nubs, these are generally things that cannot be answered with a small machine, but that should be answered before investing in a reactor-sized machine; hence, the medium-sized WB-8/8.1.

[KitemanSA]

How or from where did you arrive at the ~500W?

By other posted analyses, WB-6 = 0.6mW.

WB8 has 8 x B field and 5/3 x size.
.0006 X 8^4 x (5/3)^3 = 11W.
Dang, how did you get 500W?

If I read it right, he used a 3X increased starting point a size of 2X rather than (5/3)Xand a 10X fudge factor at the end.

.002 X 8^4 x (2)^3 = 66W. X10 = 660W.

Gotcha.

[D Tibbets]

5/3 radius increase. Where did you get that and please don't say from the picture on EMC2's website.

Well if you didn't want the answer, why ask the question?

I have looked long and hard at that image, I've even used pixel calipers on it and I just don't see any way to get valid information. I know Tom said that the 30 hole ring seal was 14 inches (I thought) but that doesn't help when there is no view point to the drawing. Or maybe it does. How? Please spell it out for me.

Assuming that the 14 inches (~35.5 cm) is the bore of the port and not the distance between the bolt holes (bore would then be perhaps several cm less), and that the design consists of individual magnetic donuts that are mounted on supports that have to slide through the port, then the upper limit on the grid diameter would be 35.499999 cm or about 34.99cm / 30 cm, or ~ 1.17 X larger. 1.17 cubed would reflect a power gain ~ 1.6X based on diameter alone. It says nothing about the minimum size or the size if the assumptions are not correct.

Dan Tibbets

[Aero]

Well if you didn't want the answer, why ask the question?

I have looked long and hard at that image, I've even used pixel calipers on it and I just don't see any way to get valid information. I know Tom said that the 30 hole ring seal was 14 inches (I thought) but that doesn't help when there is no view point to the drawing. Or maybe it does. How? Please spell it out for me.

Assuming that the 14 inches (~35.5 cm) is the bore of the port and not the distance between the bolt holes (bore would then be perhaps several cm less), and that the design consists of individual magnetic donuts that are mounted on supports that have to slide through the port, then the upper limit on the grid diameter would be 35.499999 cm or about 34.99cm / 30 cm, or ~ 1.17 X larger. 1.17 cubed would reflect a power gain ~ 1.6X based on diameter alone. It says nothing about the minimum size or the size if the assumptions are not correct.

Dan Tibbets

Ok, Thanks. I can understand that logic. It does seem to follow however (to me) that the most logical diameter would be 30 cm, unchanged from before. But, as the higher power coils will probably be thicker, a minor change in diameter should be required to maintain the central radius at 15 cm as before. With a central radius of 15 cm, and coil thickness 20% of that, or 3 cm, that gives an outer diameter of 33 cm. But the picture shows the assembled Magird with the coil thickness extending beyond the coil diameter in all dimensions. That is, 33 cm + 3 cm + 3 cm = 39 cm wide. What does that give for a resultant diagonal of the Magrid? sqrt(39^2 + 39^2) gives a dimension that is to big to fit through the opening as an assembly. Does it make sense to assemble the Magrid inside the chamber? Or does it make sense to weld the chamber together around the assembled Magrid? Or does it makes more sense that the picture is only a graphic illustration of the objective result?

@Dan - If the picture has any meaning then I accept that your analysis is correct as to assembly technique. But I think the maximum increase in the central radius could only be about 1 cm, due to coil thickness. That means (16/15)^3, or about 1.2 X gain based on diameter.

[TallDave]

Kite,

A lot of fudge in there; the 10x covers voltage and ion focussing. There's so much we don't know -- the actual starting power from WB-7, ion focusing, voltage, trapping factor, upscattering, etc. I could be off by a lot.

[TallDave]

Here was Tom on size:

On the scale of WB-8, I count thirty bolt holes on the large flanges. According to the tech data at Insulator Seal, there are two flanges with 30 bolt holes. Their outside diameters are 13.25 and 14 inches, which pins down the size of the apparatus considerably. That would rough the magnet diameters at around half a meter, so maybe they were intended as 1/4 meter radius?

viewtopic.php?p=37319&highlight=seal#37319

a minor change in diameter should be required to maintain the central radius at 15 cm as before.

I've never been sure exactly what was meant by radius in the context of r^3 power scaling. I've always assumed it was measured from center bore of the coils, because that's where the field starts (i.e, ignore thickness).

[Aero]

Here was Tom on size:

On the scale of WB-8, I count thirty bolt holes on the large flanges. According to the tech data at Insulator Seal, there are two flanges with 30 bolt holes. Their outside diameters are 13.25 and 14 inches, which pins down the size of the apparatus considerably. That would rough the magnet diameters at around half a meter, so maybe they were intended as 1/4 meter radius?

viewtopic.php?p=37319&highlight=seal#37319

I don't see that. If 14 inches is the size of the hole and the coils are to be inserted through those holes, then 14 inches = 35.56 cm is the upper limit on outside diameter of the coil. Much closer to 1/3 meter than 1/2 meter.

a minor change in diameter should be required to maintain the central radius at 15 cm as before.

I've never been sure exactly what was meant by radius in the context of r^3 power scaling. I've always assumed it was measured from center bore of the coils, because that's where the field starts (i.e, ignore thickness).

You're thinking to hard . But if you think that hard, then it becomes a very interesting question.
I have always considered the radius to describe that volume of the Polywell in which fusion reactions occur. The radius of an equivalent volume sphere, r-sub_e, so to speak. But the apparatus gives the radius of the coil of the magrid, r, as a convenient measure. So power scales with r-sub_e to the third power, and r = "Constant" * r-sub_e where the Constant is greater than one. No problem until you start changing things and try to relate the two different machines with different B field strength hence presumably different values of the above "Constant."
So now keeping my head above water means I must swim. In other words, it's to deep for me.
I am pretty sure that the 15 cm radius of the WB-6 was defined from the center bore of each coil to the center of each corresponding magnet wire bundle. In this case of fitting a physical object through a physical hole, we must be concerned with a physical measurement. The true radius of the actual Polywell is another issue. But what does happen to the Polywell radius? Does the increase in B field strength squeeze the polywell so that the radius gets smaller? It would be a pity, but I suspect that more drive power will re-inflate the Polywell ...

[KitemanSA]

5/3 radius increase. Where did you get that and please don't say from the picture on EMC2's website.

Well if you didn't want the answer, why ask the question?

I have looked long and hard at that image, I've even used pixel calipers on it and I just don't see any way to get valid information. I know Tom said that the 30 hole ring seal was 14 inches (I thought) but that doesn't help when there is no view point to the drawing. Or maybe it does. How? Please spell it out for me.

First, determine where the PoV of the graphic is. If you measure the upper and lower coils on the slant, call it ~inside 7:00 to outside 2:00, you will see that the coils are about the same size so the PoV is nearly infinite.
Next, measure the inside to inside of the port with 30 bolts and set that to ~13.25". Then measure the top coil (~22") and the left hand side coil (~24") across the total width, you will get something near 55 to 60cm. I got 50 cm the first time by using the wrong conversion ratio d'oh!!!

So, I guess my new opinion is that WB8 is 60 cm, e.g. 2X, not the 5/3X I had estimated before.

Ok?

[KitemanSA]

I don't see that. If 14 inches is the size of the hole and the coils are to be inserted through those holes, then 14 inches = 35.56 cm is the upper limit on outside diameter of the coil. Much closer to 1/3 meter than 1/2 meter.

Please folks, look at the graphic, count the holes.
On each side of the chamber there is a very large cover-plate into which is set a large flange and 4 slightly slanted smaller flanges. The large cover plate has 48 bolts. The large flange set into the cover plate has 30 bolts. That 30 hole flange is your starting point. The hole covered by the large cover plate is more like 78-80cm. It would be EASY to fit a 60cm coil thru the hole ~80cm across, no?

[KitemanSA]

I am pretty sure that the 15 cm radius of the WB-6 was defined from the center bore of each coil to the center of each corresponding magnet wire bundle.

This is incorrect. The outside diamer was ~30 cm and the inside diameter was ~20cm. Thus the major radius was 12.5cm and the minor radius was 2.5 cm.

[D Tibbets]

I am pretty sure that the 15 cm radius of the WB-6 was defined from the center bore of each coil to the center of each corresponding magnet wire bundle.

This is incorrect. The outside diamer was ~30 cm and the inside diameter was ~20cm. Thus the major radius was 12.5cm and the minor radius was 2.5 cm.

According to the final WB6 report, the magnetic coil cans were 307 mm in diameter, the inside diameter of the cans was 201 mm. Because of the spacing between the cans the overal magrid diameter was 393 mm. This would give the cans a ~53 mm thickness (17% of the overall can diameter).
By contrast, by measuring a picture of WB4, I estimated that the can thickness was ~ 25% of the total can diameter.

Why WB6 had rounded dimensions was explained, but I haven't seen why the can thickness was reduced. Keeping the same ratio would have allowed more windings, thus greater field strength and more proximity to the central face cusps. Presumably the center of the magnetic fields would have been further away from the corners, but the increased overall field strenth would more than make up the difference(?).

Dan Tibbets

[KitemanSA]

Why WB6 had rounded dimensions was explained, but I haven't seen why the can thickness was reduced.

Maybe it was simply that they had to make a unit within a certain budget and that was what they could afford.
Then again, maybe they actually used their 1.5D code to analyze the minimum size needed to make detectable fusion. After all, they had gotten fusion before, but at a massive electron loss rate. Wouldn't they know the size that would be needed?