Four Billion Degree Electrons!

[asymmetric_implosion]

Why stop at a billion? What distribution are you assuming to calculate your 1 in 14,000 chance of having 5 shots in a row that are within 3%? I think you will find that literature suggests a non-normal (non-Gaussian) distribution of radiation yield from a plasma focus. My concern is you don’t know the distribution of your radiation yield yet so claiming that 5 shots is statistically significant seems premature. In my way of thinking (engineer not physicist), each time you fire a shot you are sample the machine’s inherent distribution at a fixed set of conditions. You need to sample the distribution many times before you have confidence in it.

I’m not trying to say that the result is not exciting, but it is not unique. I appreciate that the typical standard deviation in most PF devices is ~ 50% (my PF is 46-53% depending on conditions) and you have demonstrated far better than most, but not all. Ahmed et al (doi:10.1088/0741-3335/48/6/003) showed 5% a few years ago in 10 shots using alpha particle pre-ionization. I believe your combination of the cathode near insulator and initial magnetic field are the reason for this. It is something that many others have not adopted yet.

I also appreciate the problems you’ve had with the pulse power but a reliable 1 MA drive that is demonstrated to operate at 1 Hz all day long at 100 kV costs about $5M. The electrodes and other components can’t cost more than $1M. Look up the linear transformer driver papers by Sandia National Lab. The LTD has rise times up to 800 ns. It seems that would go a long way to solving your pulse power problems in the next few years. Just my opinion but $6M seems better than $30M. Best of luck in the upcoming campaigns. I look forward to a paper on the statistical significance of the data later this year.

[zapkitty]

Why a Sandia-class power system if what they’re building up from scratch works for their particular needs at a fraction of the cost?

[asymmetric_implosion]

I was referring to the next step. Lerner stated that $30-50M would be required for the next step. My comments are directed at that statement. You can build an LTD DPF for $6M at the 1 MA level for the same pulse power and it’s highly reliable compared to the existing pulse power firing thousands of shots per day. The reports indicated that problems with the switches have caused a substantial delay. Why reinvent the wheel? The price of the LTD drops if you reduce the repetition rate from 1 Hz to say 1 shot per minute which might be enough for the next step. At 1 shot per minute you can use conventional pulse power as well but given the problems encountered thus far, why not buy a product that is engineered and tested vs. building your own? Both systems are capable of the same current, inductance and rise time (key factors for a plasma focus). If the custom pulse power can be built for less, that would be great but the problems to date suggest less expensive components were used and they paid the price for a long time patching the cheap components together and making them work. It’s always a compromise in research to purchase the proper quality of components because labor typically costs far more than materials in the end. Time wasted patching lower grade components tends to be far more expensive than buying the right part in the first place. I believe the saying is penny wise, pound foolish. Each group has to decide on a strategy that works for them but I think most companies find that buying something that works “out of the box” is worth the cost in labor. Universities rely on cheap grad student labor so they can get by using cheap components because the labor is inexpensive.

[Rezwan]

From https://focusfusion.org/index.php/site/article/562

We found that no significant difference in the ratio of the NTF to FTF signals occurred because of the filter. The x-rays were acting as if the 3 mm copper was transparent. 

This could only be the case if the x-rays were produced by electrons that had an average energy of at least 400 keV. We will keen to analyze in even greater detail results relative to the electron temperature, which is crucial for in turn heating the ions and thus overall fusion yield.

[Author]

Posts