Mad Magnets

20 02 2009

I had no idea permanent magnets could be so strong… then I saw this insane magnet accident.

Polishing the Lids

20 02 2009

Taking a page from jewelry fabrication, I wet-sanded the face of the lids on 230 grit silicon carbide sandpaper on a glass table using a circular motion. This gives you a very flat surface.


Now they fit together tightly for welding:



Flower II

20 02 2009

Another nice projection of the superconducting coils:



20 02 2009

The parts from ProMetal just arrived. These parts feel heavy, dense, and strong. You really have to hold these in your hands to believe it. They hold water without leaking. Cost ~$30 each. img_3245img_3249

There is some texture, but overall these parts are highly conformal to the design. img_3264


This is very exciting. The first prototype core is within reach.

Next I will grind the touching surfaces of the lids and try to laser weld them together.


On the down side this material is magnetic:


I wonder if ProMetal can adjust the composition of the alloy. Stainless steel can be either magnetic or non magnetic depending on the alloy:

There are different types of stainless steels: when nickel is added, for instance, the austenite structure of iron is stabilized. This crystal structure makes such steels non-magnetic and less brittle at low temperatures.

Update: Looking at the ProMetal Materials Spec Sheet, it looks like they offer 316SS, which is non magnetic, although less strong than the 420SS. Bingo.

Laser Safety Goggles

19 02 2009

For the record those OD3 @808nm laser safety googles are INSUFFICIENT for the power of this laser. I’ve ordered these glasses, which are OD 6+ @ 690-1300 nm, and are OD13 @808nm specifically. This is 10,000,000 X the protection of the OD3 glasses. They are also glass which can withstand a direct hit much better than polycarbonate. Shout out to my peeps.insufficient

Laser Welding

16 02 2009

So the next move is to attempt to laser weld two pieces of steel. Having computer controlled laser welding at our disposal will be indispensable for constructing the core (beginning with welding the lids to the chassis) and the vacuum chamber (welding the chamber itself, welding various feedthroughs). I think we have enough wattage to laser weld stainless steel, although at a slow rate.  However, we still need to focus the beam, which is especially challenging for diode lasers that produce a bar beam like this:

beam_shapeIn my research, I found the back story to the lasers I picked up. Looking at the beam path illustration, these units correct for some of the diode array’s problems with divergence and astigmatism, so the output of the diode is a rectangular beam that focuses to a 1.5 cm long line about 15 cm beyond the output aperture.  So we are working with a well columnated beam, but it’s rectangle, not a circle. 

I think the easiest route to a top hat beam would be to focus this bar beam into a single thick core multimode fiber optic cable which results in a perfectly circular Gaussian spot, and an easy way to deliver the beam to it’s target where it would pass through a simple circular focusing lens. Challenges include sourcing the correct fiber optic cheaply, cleaving the fiber optic correctly, and aligning and focusing the bar beam into the fiber optic. This process would benefit from a Laser_beam_profiler, so you can find the sweet spot with focus and alignment for max transmission through the fiber. 

Here is a good overview of the basics of laser welding, including weldability of various materials.

Fire the Laser!

13 02 2009

Today we will fire the danger laser! I have the plumbing for the water cooling working. For this we take a field trip to my friend Stuart’s shop. He has 240V mains and a variable power supply. Here are the lasers:


The labels are a bit ambiguous for the electrical connection. Is the chassis the anode, and both leads on the top the cathode? I think this must be the case. Poking around with an electrical multimeter, I find there is almost no resistance between the two terminals on top. But when you test between the either of the terminals and the chassis you get a different resistance depending on the direction of the test, which is from the diode’s electrical bias.



We will be testing these current limiting power supplies (originally from NASA)


Here is the full setup.


Bucket of distilled water for the heat exchanger (we must have run this at too high a voltage, you can see it melted a little):img_3234

And the verdict? Success! Here is a video of the setup, and a video of the trial run.img_3236

Chassis Fabrication

12 02 2009

I’ve found a vendor that can actually fabricate the chassis. ProMetal has a process that I would describe as the Zcorp process, but working directly with metal. They say they can fabricate the chassis and lids in a stainless steel / bronze alloy which has good physical strength.  Their max build envelope is 200 mm^3. This would allow for the fabrication of a dodecahedral core about the size of a basketball. The quote they gave me is within my budget. This is the only executable chassis option I have so far, so I am fixing my design on a radius of 100mm. I spoke with Prometal and they mentioned they have a larger format machine, but it’s down for maintenance  currently. 

I went ahead and ordered two of the lids from the copper coil version of the core. This will give me a feel for the process and material. I will attempt to laser weld the two lids together. lids

Chassis Redesign

9 02 2009

I’m redesigning the chassis to work with the superconducting cable:



The coils are now an elliptical torus, to better conform to the flat tape of the superconducting coil. The superconducting tape fits in the thin groove. Above the groove is a toroidal cooling channel. At the intersection of each joint is a spherical  cavity so there is room so twist the superconducting cable to pass through the joint.


9 02 2009


%d bloggers like this: