Hacker News new | past | comments | ask | show | jobs | submit login
Tungsten for radiation shielding use (prusa3d.com)
52 points by noyesno 11 days ago | hide | past | favorite | 64 comments

Interesting! I'm designing the flight computer/electronics of a 2U cubesat for academia, and something like this might work really well for our application. We were considering borated polyethylene but the cost for that is prohibitive. This is perhaps not as efficient in terms of mass-to-protection ratio, but we have plenty more room in our mass budget instead of actual budget. I also like the idea that we may be able to mold this material around particularly sensitive elements of our system like the memory devices, to which neutrons are the biggest hazard.

I'm pretty sure Tungsten and borated polyethylene are best at shielding against different types of radiation. Dense materials like tungsten and lead are ideal for blocking gamma rays and x-rays, while borated polyethylene is better at blocking neutrons.

Depending on your environment and the sensitivities of what you're shielding, you might need one or the other, and possibly both.

Disclaimer: I'm not a nuclear physicist. I learned this while researching for a hard-sci-fi novel I'm working on.

Academia? Cubesat projects?

Maybe get the kids to look into using Magnetorquers to get some spin on and half shield with this product .. and you now have a platform for directional gamma ray detection | mapping.

( Gamma counts coming from "over there" will decrease whenever "over there" is masked by the partial shielding )

There are other apps, but that one springs to mind.

Great idea! We do have 3 DoF magnetorquers on board, but I feel like the difference between shield on and shield off might need a very sensitive and directional detector..

It's a timing thing - Australia used a "we built it ourselves" detector with shielding that rotated about the crystal pack to find that missing mining source that hit the news last month.

The detector part (doped crystals + scintillation sensors) is fundamentally undirectional - those pesky gamma come in from any direction.

One side has to be shielded to bias the reading (use Tungsten perhaps) .. and by rotating the entire satellite the shielding direction changes - you now have a bulk data stats analysis challenge, do particular orientations align with counts (at various energy levels) rising or falling.

As alway, have a fiddle with a ground based setup first.

>As alway, have a fiddle with a ground based setup first.

now see, if we do that, then you're ruining the one chance I have of being the exact right person to go to space to fix the thing.

I feel obliged to mention coded apertures here too! https://en.m.wikipedia.org/wiki/Coded_aperture

for a cubesat saving a few bucks on machining probably isn't a good tradeoff

Tungsten has the highest melting point of any pure metal, I believe, as well as being close to the highest atomic number among the non-radioactive metals. It's also reasonably workable as a material, isn't terribly toxic, and has a reasonable availability and price.

The other high-melting metals near it in the periodic table all fall down on one or more of those properties. Osmium, for instance, is rather expensive, mined in only very small quantities, and reacts with air to form highly-toxic osmium tetroxide.

I think it's worth highlighting just how horribly toxic osmium tetroxide is -- it's used as a stain for electron microscopy as it is dense and binds irreversibly to lipid membranes. Quoting from Wikipedia:

> In the staining of the plasma membrane, osmium(VIII) oxide binds phospholipid head regions, thus creating contrast with the neighbouring protoplasm (cytoplasm). Additionally, osmium(VIII) oxide is also used for fixing biological samples in conjunction with HgCl2. Its rapid killing abilities are used to quickly kill live specimens such as protozoa. OsO4 stabilizes many proteins by transforming them into gels without destroying structural features. Tissue proteins that are stabilized by OsO4 are not coagulated by alcohols during dehydration.

> OsO4 will irreversibly stain the human cornea, which can lead to blindness. The permissible exposure limit for osmium(VIII) oxide (8 hour time-weighted average) is 2 µg/m3. Osmium(VIII) oxide can penetrate plastics and food packaging, and therefore must be stored in glass under refrigeration.

Worse, osmium tetroxide is volatile and sublimes at room temperature. I was told by my lab instructor that it is said to smell sweet, but that if you ever inhale enough to be able to smell it, you will have already absorbed a lethal dose.

lots of historical chemists smelled osmium and survived; that's how it got its name

Note that it's 75% by weight, not volume, which is about 15% by volume. A little disappointing.

Kind of makes sense though. Tungsten is quite dense. It can probably get prohibitively expensive pretty fast. Not to mention difficult to flow the material / keep hot the more tungsten you add.

75% by weight sounds like a lot of tungsten...

But it is only 15% by volume... So to the casual observer, this material is mostly plastic.

But its more than 3 times as dense as PETG still. 1kg of PETG makes is 300+ meter of filament, while 1kg of this stuff is only 100 meter.

So I think the casual observer would notice that. But also means with thar price per KG.

Ah, that explains why its radiation blocking is so much worse tham pure tungsten!

If you could manufacture Tungsten powder with spherical particles (rather than sharp pointed particles), I bet they could have far more tungsten in there and still have it 3d print effectively. Perhaps up to 50% by volume (95% by weight)

This material has radiation shielding properties far worse than just lead - lead which is cheaper and easier to manufacture

Pure tungsten on the other hand is very good and can even shield beta particles

Yeah but it's lead! Lead isn't a very friendly thing to just have around.

paint it, sheesh

Yeah, but you'll never pass RoHS, try shipping a product that doesn't meet RoHS...

are you saying this filament is nothing more than a meaningless exercise in box-checking bureaucracy

that is depressing

No, im saying lead is bad and you can't put it in products anymore.

Tungsten is ok

lead metal is safer than tungsten to have in your body


don't confuse government power with wisdom; in many ways they are opposites

this sounds more entertaining for 3-d printing things that startle people by being staggeringly heavy rather than things that are actually useful

it should be heavier than lead; lead is only 11.3 g/cc, tungsten is 19.3, 75% tungsten would be 14.5, but then the other 25% is petg with sp.gr. ≈ 1 so you should actually get a density of 14.7

but i struggle to imagine cases where using tungsten is a more cost-effective option than using lead and making the object 9% bigger? they cite 'various medical applications' but tungsten isn't exactly ideal for permanent body contact either

oh actually they say it's 75% tungsten by mass, not volume, so it's only 4 g/cc, and so its attenuation (at 140keV) is only 18% of lead's (by volume)

copper might be an alternative that is less toxic than lead and less expensive than tungsten

If it's only 75% tungsten, I'm figuring it's a tungsten powder. I work with lead all the time but I would not fuck with lead dust. Lead sulfide would be much safer, although studies feeding it to rats did show elevated lead levels.

That said, no metal dust is very nice. Solid lead might be safer than tungsten powder. Maybe you could make an iodinated polymer or use a barium cement.

my candidate list of high-density fillers from dernocua (sorry for markdown) is

- Osmium: [US$13000/kg][8], 22.65 g/cc, or possibly iridium at more than twice that price

- Tungsten: [US$30/kg][9], 19.3 g/cc

- Tungsten carbide? Not sure what it costs but its density is 15.6 g/cc.

- Lead scrap: 95¢/kg, 11.3 g/cc

- Steel scrap: [21¢/kg][10], 7.9 g/cc

- Magnetite: [10¢/kg][11] [or so][12], 5.2 g/cc

- Quartz (as construction sand): 3¢/kg, 2.6 g/cc

- Water: [.06¢/kg][22] or so, 1 g/cc

[8]: https://www.metalary.com/osmium-price/ [9]: https://www.metalary.com/tungsten-price/ [10]: https://www.usgs.gov/centers/nmic/iron-and-steel-scrap-stati... [11]: https://www.usgs.gov/centers/nmic/iron-ore-statistics-and-in... [12]: https://stockhead.com.au/resources/barry-fitzgerald-why-magn... [22]: http://www.scientificamerican.com/article/israel-proves-the-...

this is my approximation of the pareto frontier; that is, each of the items on the list is conjectured to be cheaper than everything that's denser than it is, and denser than everything that's cheaper. corrections are welcome

i was thinking baryta (46¢/kg, 4.48 g/cc), mercury, litharge, minium, cinnabar, cupric oxide (US$3.90/kg, 6.315 g/cc), zinc oxide (US$29/kg, 5.6 g/cc), and manganese dioxide (5.026 g/cc) might be interesting in this context too

i hadn't thought of your suggestion of galena (just cinnabar) and generally i'm skeptical of metal sulfides because of their tendency to produce hydrogen sulfide; i don't think that's an issue with those two. litharge, minium, and mercury are a lot more worrisome toxicologically

i don't have solid pricing information for mercury, litharge, minium, cinnabar, or galena, and i'd be interested

tungsten is probably more chemically inert for medical purposes than a lot of these

Density alone is a weak parameter. For photons with energy below 200 keV — that is, most of them — the attenuation is proportional to Z^2, so barium with atomic number 56 is four times as attenuating as nickel with atomic number 28, and lead with atomic number 82 is about twice as effective as barium — all of this before you account for density.

I've done a bit of searching for materials myself. Barium is mined as barite (BaSO4) or witherite (BaCO3), not baryta (BaO*xH2O, caustic), and USGS lists the price of barite as $180/t, or $0.20/kg.

You also have a K-edge effect, which prompted me to wonder whether you can easily produce barium zirconate from the respective ores, which are both cheap — BaZrO3 (sg ~5.5) is not currently manufactured (Zircon sand was <$1/kg until a COVID-related shortage). But at this point I decided I was overthinking it.

baryta is barite, not baria. baria is not just caustic but also toxic and prone to produce dangerous peroxides

i think the price i cited for it is retail, locally

i was interested in density for psychological effects when i made the list

thanks for the tip about barium zirconate; it sounds very promising. how about lead zirconate?

I'd think the public palatability of lead products is one aspect. People don't worry as much about tungsten.

I was thinking weights to fit specific items. Model railway wagons, for example, need to be weighted and balanced to operate smoothly, but you have significant size constraints. I threw a bunch of tungsten weights inside a small locomotive to give it extra tractive effort.

yeah but it turns out this is only 4 g/cc

Might be interesting to print grip modules for firearms (Sig does the "TXG" series which are tungsten-infused polymer, used for adding mass within the defined shape/volume for competitive shooting/recoil reduction reasons. Due to the design of the firearm, the grip module is just an accessory and not a firearm and thus not regulated, at least under US law. Being able to print custom shapes would be cool. (https://www.sigsauer.com/p320-x-series-txg-grip-module-assem...)

The density is only 4 g/cm³.

You'd likely do better just machining it out of steel (5g/cm³, and much cheaper)

The 3D printers which can do metal (metal forge, etc) are incredibly expensive (100k+). I was thinking about custom to hand 3D printed grip modules, qty 1. Machining a grip module out of metal requires tools and skills beyond 3D printing, too. There are also some slight performance advantages of a polymer frame (grip module, in this case) over a metal frame.

That is actually so useful for me, wow. We prototype in lead, which is a pain from an OHS perspective. This would make life a lot eaiser, without the cost of tooling to get moulding nylon/tungsten parts made.

- "The calculated HVL for Prusament PETG Tungsten 75% is 1.402 mm (orange mark). For comparison, the HVL for pure lead is 0.256 mm, HVL for pure tungsten is 0.191 mm."

Err, they make it sound like ordinary lead is the best choice? The article's point is that additive manufacturing is a workaround for tungsten's difficult material working properties. But: lead foil, you can simply bend it with your hands, into any shape you want. And apparently it's much thinner.

In other words: you need less tungsten than lead to get similar shielding results.

But the tungsten costs many times more, and is also much harder unlike lead which is soft and easily worked, which is why radiation shielding is still overwhelmingly made of lead. In applications where its toxicity is a problem, it's used encapsulated inside another inert material.

Sure, if you want to shield a wall or container or something you will just slap lead on it, but this could be used for complex shielding of sensitive components in nuclear machinery and devices. Machining hard materials is extremely expensive and difficult (exotic material cutters or edm, with slow high precision grinding).

Lead shielding is usually applied in sheets; and due to its low melting point, it can also be cast as a layer on other machined parts on applications where sheets are more difficult to use.

so machine a die, machine your part, build random fixtures, cast lead on to your part...

or just print it?

It's a bit of a no brainer.

Even if you just want a shielding layer, the convenience of this will win out on low volume parts.

lead isn't a hard material, and if you need a hard material, this filament isn't what you're looking for either; it will be almost as fragile as pure petg

I didn't say it was hard. I said machining tungsten is hard. Machining lead is also difficult and weird and the end result can be barely usable for anything, especially in hot environments or with any load. It's just not good.

If you need complex parts, this could be an excellent choice.

Idk what the confusion here is. Maybe you are unfamiliar with machining?

The first sentence of your comment was structured like this: [reason to use lead], but [reason to use tungsten]. At least I expected the second sentence to expand on the pros of tungsten, when in fact it was about the pros of lead; interpreting it as the former would mean lead is a hard material.

The article explains the beneficial properties of tungsten with rad shielding. Tungsten sheilds better than lead. Lead is easier to form to simple shapes (flats, straight forward wraps). It is essentially impossible to form it in to complex shapes. Tungsten has a high temperature resistance, and low deformation under load. Broadly speaking it can be relied upon to stay in the shape it's machined to. But getting in to that shape is very expensive and difficult - an understatement.

I can't explain everything in every comment. You can research the materials if you want to understand the discussion in more context.

Tbh I have no idea how you came away thinking I said lead is hard. "also difficult". "this" is the material that is the center point of the discussion. Context, yo.

> Tbh I have no idea how you came away thinking I said lead is hard.

I did not, in fact, come away thinking lead is hard. I did, however, explain and spell out for you how @kragen probably did so; I'm not sure how I can make it any more clear.

i appreciate you being willing to clarify the things that were hard to understand or that we were wrong about

(i actually had the impression that working lead on a lathe was very easy indeed due to its softness, but i've never tried doing it myself, and my bachelor's degree from youtube is worth what i paid for it)

i think you may have misunderstood something in the comment you were replying to yourself

Lead is a nightmare to machine. It's gummy. You'd think aluminum was easy because it's softer than steels, but that comes with it's own challenges. Lead is worse so. You're right that this isn't so bad on a lathe as opposed to a mill or (lol) a grinder, but you're not going to see a need for a lot of solid round shielding.

A little bit of lead in steel will increase machinability, but only a teeny amount. Similar to adding a small bit of phosphorus.

It's also hard to hit dimensions in lead, and if you do hit them, the second the temperature changes you'll lose them. Additionally, whatever you make can't see any stress, or the part is donezo.

Idk if you've ever handled lead - but considering you can bend thick sheets of it by hand and melt it on your stove, I'm sure you can imagine the kind of issues you might have integrating expensively machined chunks of it in to hot high pressure environments with moving parts.

Even though lead is cheap and this filament is expensive, actually getting the lead to shape by machining or working it is also a costly process. 3d printing has a sort of cost ceiling. Once a part is designed, all of the real work is done. For fabrication and machining - once the part is designed the work has only begun.

What might look like a small and simple combination of geometric shapes can cost thousands to machine.

I'd rather spend 500 on filament and a day's engineering labor on a part than the same day's engineering, 100$ of lead, and 1k or more on manufacturing. Not to mention the lead times on machined parts can be wild.

i think all the post-manufacturing problems you describe with lead — dimensional instability, thermal creep, plastic deformation, incompatibility with hot high pressure environments, vulnerability to wear — are even worse with this filament

maybe an exception is that petg's maximum elastic strain is larger than lead's

Considering lead's modulus of elasticity is essentially zero, yes :) You can drill, tap, and generally rely on this material to stay in the shape it was printed at in maaaaaany more situations than lead.

Of course the petg in there is a limiting factor, but even plain petg is an order of magnitude better than lead at retaining its shape under a wide range of loads. You can print gears for lathes in petg! And they last years! (If you're wondering why you might do this - it's a good idea to have a cheap point of failure on devices that have enough power to rip themselves to shreads.)

Not to mention this is only v1 of the material. We could see exotic plastics that are far more heat resistant like PEEK get a secondary filler for niche radioactive usecases.

https://www.matweb.com/search/DataSheet.aspx?MatGUID=ebd6d2c... says lead's modulus of elasticity is 14 gigapascals and https://www.matweb.com/search/datasheet.aspx?MatGUID=cb9580a... lists one variety of petg as having a modulus of elasticity of 2.2 gigapascals; i think you are confusing your material properties with one another

(at such low filler loadings the tungsten will change the modulus of elasticity very little)

petg does have higher yield stress than pure lead (53 megapascals vs. what https://nickelinstitute.org/media/1771/propertiesofsomemetal... says is 17 megapascals) but there are lead alloys that are in the same range of yield stress, including regular lead-tin solder. but they won't last if you try to make change gears out of them precisely because they're harder than petg

by the same token, though, i think you're going to get a lot of springback and long-term distortion out of petg you aren't going to get with lead

well i'm certainly no master machinist, heh

you seemed to be saying that this filament might be a better alternative to lead sometimes, but i couldn't tell when that might be. pure tungsten is clearly a better choice sometimes, for example because it shields better than lead, is harder, is denser, and is more refractory, but none of those seems to be true of this filament

from your other comment at https://news.ycombinator.com/item?id=35206874 it looks like you're saying that, although this composite is inferior to lead in those ways, it's easier to print

We have issues working with lead at work, mostly due to toxicity. Putting a hole in a sheet isn't as simple as getting out the drill, not mechanically, but due to all the swarf that needs to be cleaned up. 3D printing a bolt-on cover would be lower risk than drilling lead, a quicker turnaround than getting a die made and lead cast, and give a prototype that doesn't need to be handled with gloves.

For the right application, there are some good wins here.

i appreciate the useful feedback

what do you think about copper-filled or baryta-filled filaments as cheaper alternatives

maybe this is wrong but i feel like both tungsten and copper are in the 'if you have it embedded in your body you are going to need surgery to get it out before you get gangrene' while lead and baryta are not

I bought a 1kg tungsten cube

Top-quality paperweight.

Neat. If I was in dentistry or oncology I'd print a tungsten apron like a suit of armour for kids, or perhaps a nice comfy/adjustable thyroid collar which isn't too tight or tall. With filament at $230 a kilo it might not be economically feasible but it sure would look cool.

Perhaps 3d-printed tungsten filament chainmail?

I printed a hat. Works great for the silent radio waves that the government uses to control your mind.

Don't let this troll fool you. Tin-foil hats only work when they're made of tin; not tungsten, not aluminium or aluminum, only tin.

They also need to be grounded, if they're only in contact with you they're an antenna.

Foiled again!

Good neck workout too.

Applications are open for YC Summer 2023

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact