Serious HES derivative project proposal

[r2fpl]

You're right. Sometimes one detail changes everything or many do nothing. One difference is the reception method which is different. Maybe it's this detail, but I'd have to run the whole device to check it. I wish it worked but I know how much time I wasted. I know dozens of such projects and none of them is true in the good results.
For now, I do not know what would convince me to return to this concept of the device. Maybe someone wants to do it too.

[Cadman]

Hi guys,

This is definitely a better way way to make the laminated cores than what I started to do.

Much thinner PLA separators between the iron layers and between the core and coil wire. The separating layers are down to 0.013" and the thickness of the mold casing is right at 0.020".

The idea is to put down a single layer at a time followed by a separator, compress and cure ... repeat.

It takes about 4 hours to print a mold section for the Samsung rotor core but you only need 4 of them for this. The separating layers take right at 12 minutes each and don't have to be trimmed or anything. Ready to use as soon as they cool.

In one of the images below the light blue-green is on separator sitting at the top of the core mold.

Much easier.

[SolarLab]

You're right. Sometimes one detail changes everything or many do nothing. One difference is the reception method which is different. Maybe it's this detail, but I'd have to run the whole device to check it. I wish it worked but I know how much time I wasted. I know dozens of such projects and none of them is true in the good results.
For now, I do not know what would convince me to return to this concept of the device. Maybe someone wants to do it too.

r2fpl,

Sorry to hear you are giving up on this device.

Should you ever decide to return to this concept of the device using an Analytical (mathematical)
design and analysis or a Numerical approach I've attached a pdf file that you may find of interest.

Sometimes a little up-front engineering work can save a lot of time and expense. Having analyzed
dozens of so called FE discoveries; this is only one of two that showed great promise and has actually
worked out so far.

Anyway; the first recommendation is to find access to a professional CAE Suite (school couse, etc.),
or try an Analytical analysis (using your favourite math program, or whatever). The attached file
might provide some insight and guidance as you engineer the solution. 

The first five chapters are generic with respect to developing most devices of similar structure,
including the LinGen. The remaining chapters are specific to designing a "Fault Current Limiter" but
still contain some useful magnetic circuit information.

Both Analytics (Math) and Numbeics (CAE) are covered. Modern CAE has developed quite a bit with
the addition of GUI's, etc. since this writing. Taken from notes of an old COMSOL course.

It's quite comprehensive and well done IMHO. Provides some good reference material as well.

This concludes my contributions and presentations regarding the analysis, development and design of
a specific embodiment found in one of the Holcomb patents. 
SL

[Cadman]

Just a quick update.

Casting the laminates has been problematic so far. The polymer I have used for smaller laminates is not strong enough for these large ones with such narrow sections and they are cracking when even slightly flexed.

I'm switching to a two part fiberglass resin now and hopefully that will solve the problem.

[Cadman]

Laminate experiments continue. Up 'til now I've only managed to keep from reducing the density too much below the 2.52 g/cm³ apparent density of the loose powder.

I may have found the ideal composite powder for DIY though, ATOMET EM-1. If I can purchase small quantities and also figure out an economical way to compress it to about 420 MPa. That pressure would give it a density about 6.8 g/cm³ which is better than ferrite.

Its most attractive feature for DIY is that it cures in 30 minutes at 200 C. No expensive kiln needed.