How to actually mine Platinum Group Metals from an asteroid

How to actually mine Platinum Group Metals (PGM) from an asteroid, or my guess at how Planetary Resources might do it:

1) Find a metal-rich rubble pile (if they exist) with a PGM ore concentration of 100ppm (asteroids of that concentration of PGMs supposedly exist).

2) Pick off big metal chunks magnetically or somesuch.

3) After you have a thousand or so tons of metallic ore, put it in the shade to cool off to deep cryogenic temperatures.

4) Then, crush it in a grinder. Cooling the ore to cryogenic temperatures first should significantly reduce the energy needed to grind the ore to a fine powder, but grinding of ore often takes roughly ~10kWh/ton, so it'd take ~10MWh to grind the batch (or about 2 weeks if you have 30kW), although iron-nickel is pretty tough, which is why I included first cryogenically cooling the ore.

5) Grinding may heat up the ore, so it may take several steps where you grind, then cool, then grind.

6) Once the powder is fine enough, you put it in a vibratory centrifuge to clear off the relatively lightweight rocky bits (the metal will be much denser).

7) Then use a magnet to remove the powdered iron and nickle bits.

8) What you have left will be a MUCH more concentrated ore and a lot less material.

9) You can ship this back to Earth (a couple of tons to a few hundred kilograms... small enough to be returned to Earth, perhaps guided by a very small tug and perhaps with a very inexpensive heatshield) for sale or further processing.

10) Or do some wet-processing of the ore to concentrate it to PGMs (like dissolving the ore in acid... PGMs won't dissolve in most acids while everything else will, so you'll be left with fairly pure PGMs) before shipping back to Earth for sale.

However, acid processing may be too chemically intensive, and you might get by with just dry processing of the ore (although it sounds like Planetary Resources intends to do some wet processing, part of their motivation for focusing initially on extracting water).

At the end of this, you'll have about 100kg of PGMs, with a market price of roughly $15-30 million.

Rinse and repeat.

First test firing! No continuous ignition.

This is the first test of a rocket igniter (i.e. small rocket engine) running on compressed air and propane. Using compressed air as the oxidizer is notorious for causing difficulty in keeping a rocket engine lit, as you can see here where the rocket only ignites a few times (by the spark plug) and never stays lit for any length of time. Basically, it appears like the spark just ignites what's in the ignition chamber before it all goes out. I tried some nozzles slightly smaller and slightly larger, and one just very slightly smaller seems to work best. (I also tried increasing the spark frequency from 31Hz to 122Hz, which helped slightly.) All of them seem to work best after the chamber becomes warmer after more testing. I may try running it for longer to warm it up and see if it ever stays lit for an appreciable length of time. I'm just concerned about melting the plastic tubing which brings the propellants into the chamber if I get the chamber any warmer (yeah, I probably should've tried metal plumbing there...). I may try seeing if I can water-cool it a little.

http://youtube.com/7alGMFgz52s

I originally had a lot more trouble with the spark generator producing a whole bunch of EMF which caused the microcontroller to go crazy. The relay which I was using for isolation earlier wasn't helping anymore for some reason, so I switched to an LED/photoreceiver setup which worked pretty well then had its own problems. I separated the power supplies to separate AC circuits and then used a metal mesh to shield the non-sparker side of the test stand, and that seemed to do the trick, 95% of the time, at least (good enough for me, though I can do better). I think I'm slowly wearing out the microcontroller somehow.

If I'm ever able to get it working smoothly, I will make a parts list. I need suggestions about how to keep it lit.
Here's a picture of the chamber with the nozzle unscrewed:

Choked Flow! Mach Diamonds! Schlieren! (but just with compressed air)

Here are a couple "Schlieren" images I made (really just shadow graphs) using a DSLR camera without a lens, the bright video flash (just a bright white LED) on my cell phone, and compressed air going through the oxidizer orifice that will be used for my rocket igniter (I just drilled it totally free hand, no drill press... 1/16" diameter orifice, I think). You can see nice shock diamonds, especially with the last image (which was taken with less pressure drop, actually... I was running my air compressor in "blow down" mode because it's loud). I need to put my pressure gauge closer to the orifice (should be on the orifice side of my valve) to get a better reading of the pressure drop, but I think the first image is at about 100psi, with the next one being significantly less.

I simply projected the shadow of the end of the orifice directly on the DSLR image sensor. I held the orifice pretty close to the camera. To get more contrast, I could have held it further away (allowing the diffracted vs undiffracted light to diverge more, giving more contrast).

And here's what the orifice actually looks like, next to my breadboard that controls the spark generator (indirectly through a relay, ~31Hz PWM) and the valves, with three LEDs hooked up so I can test the code without the valves and stuff connected:

The purple thing is just the FTDI cable used for programming the ATMEGA chip and powering it with my laptop (in the field, it has its own 5V power supply). FTDI is optional, I'll probably remove it once everything is integrated.

I'll post more later.

Calculating vapor pressure of propane during test firing

I will be relying on the vapor pressure of a propane tank to maintain a certain feed pressure for my rocket igniter (which I will measure with a pressure gauge). But will the boiling/evaporating of the propane while it's being used lower the temperature enough that there's a significant reduction in feed pressure?

Using figures mostly from here:
http://en.wikipedia.org/wiki/Propane_(data_page)

propane heat of vaporization:
356kJ/kg (or 356J/gram)

propane liquid molar heat capacity:
98.36J/(mol*K)

propane molar mass:
44.1 gram/mol

propane liquid heat capacity:
98.36J/(mol*K)/(44.1 gram/mol)
=2.23J/(gram*Kelvin)

Thus a 4 second firing consuming .5 grams of propane a second for a total of 2 grams will cool the liquid propane in the tank by:
2grams*356J/gram=712Joules
which will change the temperature of half a liter (roughly 500 grams) of propane by:

712J/(500grams*2.23J/(gram*Kelvin))
=
.639 degree (in K or C)
that's roughly 1 degree F (or Rankine)


It's quite possible that the whole tank won't equalize in temperature, but even so, it does mean there's not a big enough drop in total temperature to lower the overall vapor pressure by a considerable amount over the length of the burn.


Here's a chart of the vapor pressure of propane (and other substances) as it changes with temperature:

Wolframalpha.com says that propane at 77 F has a vapor pressure of about 138.1 psi, and at 76 F, it's 136.1 psi. So, if we're operating the tank at 77 F, we'll get roughly a 2 psi change in feed pressure over the length of the burn, assuming 500 grams of propane in the tank, a feed rate of .5 grams of propane per second and a 4 second burn length. That's low enough that we don't care.

BTW, since we'll be wanting to operate at closer to 80psi for the propane tank, that means we need it to be at 41 Fahrenheit (5 C), which means an ice bath or something like that.
http://www.wolframalpha.com/input/?i=propane+vapor+pressure+at+5+Celsius
Still need to figure this out, but I will be doing some experiments with flow-rates and orifice sizes (I'll be drilling them my own).

Another update, now in 2012

Well, it's been a couple months since the last update. HOWEVER, I now have renewed my subscription to my local hackerspace, acquired the block of brass (for free), have very nearly acquired the coil-on-plug (for free, just have to pick it up), and have some favors I can call on for my local machinist. I also bought a decent CAD program, Alibre Design Personal, which makes certain things a lot easier (like putting threads in my drawing and making a 2D diagram).

And now, I think the valves need 12V and 24V, so that's three power supplies I need (5V for Arduino... could be from laptop, though I may want to be better isolated, 12V for one valve, and 24V for the other, and 12V for the coil-on-plug)

Still need:
*the right taps for my spark plug and my orifices (if I don't just drill them) or my existing adapters
*correct gender for the propane tank adapter
*orifices (again, if I don't drill them)
*propane tank (may have one lying around)

Semi-optional:
*filters
*usb opto-isolator (or, just run it from a cheap PC or semi-broken laptop).
*digital pressure and temperature probes

I've also been thinking about making the throat and nozzle replaceable, perhaps allowing me to 3d-print different nozzles (in steel) for testing different shapes in subscale.

Also been thinking of using two, three, or four servos to do some experiments with vectored thrust, ala V2 and Redstone, etc on a subscale. I've got 2 servos already.

I have to make sure I can safe the system by just turning the power off on everything. I'll have everything hooked up to a power strip I can just switch off or unplug everything if something goes bad. Also, I want a big, red "fire" button. ;)

Fire durations will necessarily be very short, one the order of a few seconds, starting with just two seconds.

I just got a nice digital SLR (not video-capable, though), so better pictures will be up shortly.

EDIT: Most of the stuff I got:

First snow and time for an update...

It has been... almost 7 months since my last update. There's something about summer that makes you want to put projects like this off! I have also been taking a class at the local university about space plasma physics, so that has taken up my time as well (and likely will slow me down until it's finished). Well, I have acquired both the valves (for fuel... propane... and oxidizer... compressed air), and tested them with compressed air. They both work great! They use different voltages, but that's not that bad (12V and 5V, I believe). I also have a manual ball valve, which I plan to use for the propane side for added safety.

I also acquired connectors for connecting a small, camping propane tank to my plumbing, but got the wrong gender attachment. I also have a pressure gauge, which I think works with propane as well.

What I need now is:
*the right gender propane tank attachment piece
*small propane tank
*spark generator (one of those coil-on-plug deals which just needs 12V and a TTL trigger or something like that and outputs a nice spark to my spark plug)
*a block of brass or something like that to machine the igniter out of
*perhaps some orifices to get the right flowrate for each side... possibly could just machine these directly into the brass block, but it'd be more difficult
*filters to put just upstream of the orifices to catch foreign debris and keep it from exploding (which it might anyway)
*5V and 12V power supplies with enough current for everything

Nice-to-have:

*maybe some sort of optoisolators to protect the arduino?
*maybe one or two extra, cheap arduino clones, like Sippino, so I have extras
*digital pressure sensor for the chamber, ideally pressure sensors for the chamber, one for oxidizer, one for fuel
*digital temperature probe for chamber

Eventually, I'd like to get some more cheap activated valves and make a mini cold-gas thruster system, with the little thrusters printed online. That'd be cool.

A few WEEKS of no new progress.

I have made no new progress. But I am committed to following through on this project. I am going to find some valves which can be actuated.

After 3 days of nothing done, some pressure vessel calculations, Day 13

After being busy and not being able to do much, I've done some calculations on the pressure vessel capabilities.

Brass 360 (the alloy I think I'm using) has a yield strength of about 45000 psi, and an ultimate tensile strength of 58000 psi. We assume a typical ASME factor of safety of 4. In the cross-section, the threads will be about 19.5mm in diameter, in a block of metal with a square cross-section of 1.5 inches on a side, leaving us with walls about 9.3mm thick at the thinnest. That means the wall thickness is roughly the same as the radius of the whole, meaning that, if we abuse the thin-walled pressure vessel equations, we can get roughly 10000psi internal pressure. Of course, thin-walled pressure vessel assumptions only work if the wall is roughly one tenth the radius, so in our case, we can make a very conservative estimate (remember, this is on top of the factor of safety) that 1000psi should be safe, at least in cross-section. And even that is an order of magnitude higher than we expect, so we will likely have a factor of safety of at least 100 if we carried out all the necessary calculations (and you include the ultimate strength). So this allows us to possibly survive a hard-start even if we start using liquid fuels, and also gives us some margin for changing of the properties with increased temperature.

Found out spark plug is larger than expected, Day 9

So, I cut a piece of wood roughly as big as I expect my igniter to be (1.5"x1.5"x3"), then after talking to a guy named Michael, I thought I ought to check my spark plug, and I found that it's 3 or 4 mm larger in diameter than I expected, which actually means I can make the chamber shorter (which is good and gives me more of a chance of being able to use the drill press for it).

Another thing I learned is that I probably will want to use the lathe for the rocket igniter, although I still will most likely use the drill press since I know how to use it.

Tomorrow, I will update the CAD drawing and will show a picture of the piece of wood I am going to use as a mock-up.

Also, I will calculate the pressure vessel strength of the combustion chamber. Maybe not tomorrow, though.

Nothing done, Day 8

Nothing done for the rocket igniter today, but tomorrow I hope to go to the Hack Factory again with the goal of looking around and finding some more orifices. Also, I will try to cut a block of wood to try making a better mock-up of the igniter.

Fitting pictures, Day 7

As promised, here are the pictures of the two orifices I plan on maybe using:

I may have to use different ones, but these are the two I have to work with right now (I can get more of each kind).

In the meantime, I'm learning a real CAD program, Alibre.

Propane flowrate, Day 5 & 6, a little late

For propane:

The formula for subsonic mass flow rate through an orifice is:

http://en.wikipedia.org/wiki/Orifice_plate#Calculation_of_expansion_factor

mdot is mass flowrate (kg/s)

C is the orifice flow coefficient (unitless), .61-.9 (we pick .7)

A₂ is cross-section area of orifice (m^2), ((.0265 in)^2) * pi = 1.42334131*10^-6 m^2

ρ₁ is upstream gas density (kg/m^3), ~11 kg/m^3

P₁ is upstream gas pressure (Pa or N/m^2), 80 psi = 551580 Pa

P₂ is downstream gas density (Pa or N/m^2), 60 psi = 413685 Pa

k is the ratio of specific heats (dimensionless), 1.127 (for propane at room temp)

http://www.google.com/search?q=.7*(.0265in)^2*pi*sqrt(2*11kg/m^3*80psi*(1.127/(1.127-1))*((60/80)^(2/1.127)-(60/80)^((2.127)/1.127))) = 0.00143057116  kg / s

So, the mass flowrate for propane will be about 1.4 grams per second, about the same as for the air flowrate (except we are operating the propane at a lower pressure). This means I'm going to have to have a much different orifice size for the propane versus the air, since the propane should only be about 4% of the total mixture by volume.

It's after midnight, and I didn't get to show pictures of the existing orifices for air and propane. I'll do that Monday sometime.

Orifice sizes, redo CAD drawing for right dimensions, Day 4

Well, I didn't get the chance to calculate the orifice size last night, but I have done it today.

The formula for subsonic mass flow rate through an orifice is:

http://en.wikipedia.org/wiki/Orifice_plate#Calculation_of_expansion_factor
mdot is mass flowrate (kg/s)
C is the orifice flow coefficient (unitless), .61-.9 (we pick .7)
A₂ is cross-section area of orifice (m^2), ((.0265 in)^2) * pi = 1.42334131*10^-6 m^2
ρ₁ is upstream gas density (kg/m^3), 7.2 kg/m^3
P₁ is upstream gas pressure (Pa or N/m^2), 90 psi = 620528 Pa
P₂ is downstream gas density (Pa or N/m^2), 60 psi = 413685 Pa
k is the ratio of specific heats (dimensionless), 1.4 (for air at room temp)
http://www.google.com/search?q=.7*(.0265in)^2*pi*sqrt(2*7.2kg/m^3*90psi*(1.4/(1.4-1))*((2/3)^(2/1.4)-(2/3)^((2.4)/1.4))) = 0.00137946227 kg / s
So, the mass flowrate for air will be about 1.4 grams per second. A little more than that will be the total, since we have to add propane (which is going to be ~4% by volume). I should add that this is with using one of my existing orifices for the air. I may choose to use a larger (existing) orifice for air and use this smaller orifice for propane, but honestly I'm too tired right now to calculate any further tonight.

I also realized today that my chamber will be about 16.5 mm in diameter (not 18mm), since the recommended drill size for a 18mm diameter with 1.5mm threads tap is 16.5mm (or 21/23") according to this: http://www.carbidedepot.com/formulas-tap-metric.htm

That forced me to redo my CAD drawing. I also decided to make it not as thick to go with a cuboid of 1.5"x1.5"x2.75" instead of 2"x2"x3" in order to save a little on material costs. This will allow me to use bronze for about the same price if that turns out to be easier to drill. Bronze reported works better than stainless with liquid oxygen, though of course we're nowhere near liquid oxygen right now. You can't tell much difference in this picture, though. I've also started to learn how to program OpenSCAD, which is a CAD program which doesn't have a WYSIWYG interface, but instead allows you to model with a kind of programming language. This would make it pretty easy to change parameters automatically or with a formula, which may be handy in the future, though it's sort of limited, now. Reminds me of POV-Ray, which I used to use a lot "back in the day."

This weekend, I will try to nail down the orifice size for the propane side and trade my orifice options. I'll also take pictures of the orifices and possibly (if I have time) try making a mock-up of the rocket igniter using wood (allowing me to actually use the tools I will need to use for the metal one, allowing me to check clearances, etc).

CAD drawing of simple rocket igniter, Day 3

So, here is a CAD drawing of my rocket igniter. The small grid spaces are one millimeter on a side. It will be machined out of a block of metal (most likely stainless, maybe steel or brass) using a drill press. I don't know how I'm going to do the nozzle, yet, though.

The top will be threaded for my 18 mm spark plug, and the two holes, one on each side, will be threaded for my orifices (the size of which I haven't calculated, yet... should finish that up tonight). I may also just drill the orifices directly in the metal, though I'll still have to thread holes for attaching the propellant hoses to them.

First sparks, Day 2 (continued)

Well, I went to my hackerspace TC Maker at the HackFactory (http://www.tcmaker.org/blog/hack-factory/), and got a small group of people brainstorming for me about how to drive the spark plug. Turns out there was a semi-broken neon sign in the corner. We hooked up the spark plug (in series to the rest of the sign that wasn't broken), and we got first sparks:

Here's the video:

Well, I accomplished all my goals for today!

For tomorrow: calculate orifice size and make CAD model of igniter.

L*, Characteristic Chamber Length, Day 2

One of the first things you need to do for sizing your rocket chamber is to find the characteristic chamber length (L*) for the propellant combination you've chosen. A list of values can be found here: http://www.braeunig.us/space/propuls.htm#engine
Excluding hydrogen peroxide (which needs a catalyst bed) combinations, the longest L* is 127 cm (typical values are between 50 and 100 cm, or so I've heard). Based on anecdotal searching around the internet, I get the impression that using gaseous propellants means your necessary L* is lower. I've heard values of L* around 80 cm (~31.5") being used for propane/air (though I think that's conservative), so let's go with that as a minimum.

Given a certain throat area and chamber cross section area, L* (L star) helps you find how long your chamber needs to be:
http://www.braeunig.us/space/propuls.htm#engine

L* is characteristic length
V_c is chamber volume
A_t is throat area.
So, we can re-write. that as:
L*=A_c·L_c/A_t
where: A_c is chamber cross-section area, which is fixed for us at ~pi·(18mm/2)^2 = ~2.545cm^2 = ~.3944 in^2
and
L_c is chamber length, which cannot be more than travel length of our drill press at 2.5 inches minus length of the spark plug into the chamber which is .75 inches or so, giving a max of about 1.75". Ideally, we'd want a little less than that so we can drill the nozzle straight through with the drill press, so ideally we'd want about 1.25" for L_c, though that's a little too short, as we will find out. (I still assume L_c is 1.75" in the rest of the section, here.)

We can rearrange the above to get:
A_t=A_c · L_c/L* = .3944 in^2·1.75 in/31.5 in = ~.02191 in^2
pi·r^2=area, thus 2·sqrt(area/pi)=diameter, so: 2·sqrt(.02191 in^2/pi)=.1670 inches, or a little less than 11/64" (but more than 5/32") for the throat diameter. This fits in with my previous Back-of-the-Envelope calculations, which gave me a throat diameter of about 1/8" (the older BotE calculations were based on conservative assumptions on the steady-state capability of the air compressor I have access to, which is something like 5 SCFM at 90psi).

So, I want to make my nozzle throat no wider than 11/64" diameter and use the full travel-length for my drill press. I think I'll go with 1/8" throat for a nearly 100% margin in the throat area (I can always bore it out larger). That's a pretty small rocket, but as long as it's flamey and makes mach diamonds and doesn't leave me broke or dead, I'm happy. :)

Introduction, Day 1

I am Chris. I live in Minnesota, have a B.S. degree in Physics, and am trying to introduce myself to the field of rocketry and aerospace. To do this, I am planning to first build the simplest (safest?) bipropellant rocket engine I can muster. It is really more of a rocket igniter, but the only real design goal is to achieve ignition and supersonic flow (i.e. see mach diamonds). I figured the easiest way to get started is to set the bar low enough that I can confidently expect to finish it soon and within my available budget but still gain experience with most of the systems needed for a more complicated rocket engine. I don't want to deal with liquid oxidizers right off the bat, for instance.

It will be small with thrust in the range of single digits pounds of thrust, chamber pressure of somewhere between 25 and 100 psi, will fit in a cube about 2 or 3 inches on a side, and will use gaseous propane for fuel and compressed air for oxidizer. It will not be cooled, but I expect it will probably run cooler than other engines which use a more concentrated oxidizer.

I plan to make it out of stainless steel hopefully only with a drill press as far as power tools go (I have access to better tools, but I do not yet have experience with them). It will be ignited by a sparkplug which is as large in diameter as the combustion chamber (about 18 mm). I will try to update this blog each day to try to keep making progress until I am finished. Today, I bought the spark plug, got a tap for the spark plug on order (with no commitment to buying it) and called the metal stock company to ask about stock. Tomorrow, I hope to get sparks from it.