geekhack
geekhack Projects => Making Stuff Together! => Topic started by: sinusoid on Tue, 15 November 2016, 20:11:46
-
I've seen some plastic sandblasting fails, and thought about doing a short guide on how to avoid messing your stuff up.
Also because I've seen a lot of sources saying plastics shouldn't be sandblasted. Ever. Which is not true.
I got my inspiration from a very nice topic over at http://www.mig-welding.co.uk forum, I'm compiling and reposting it on the bottom, for posterity. I've done a bit in the subject myself, so I'll get you some theory first. Feel free obliged to correct me if you see a mistake!
Sandblasting generates friction. Friction generates heat. Plastics are good thermal insulators. So even though you're spraying them with air, they are likely to build up enough heat to melt, burn, and warp if you're not careful. Protip: if you're 3d printering, avoid sandblasting PLA. It's extremely hard, impact resistant, and has a miserably low meltpoint. PLA and sandblasting is trouble.
Air itself close to the nozzle can generate enough friction to melt the plastic. It's not uncommon to get supersonic flow with a compressed air nozzle. So keep a safe distance between the nozzle and your precious plastic parts. If you want to smooth out 3d prints, use bigger grit, or do it manually.
Sharpness of the grit matters. You want an abrasive that has sharp edges, like corundum. If you'll use riversand or glass beads, you're in for a world of pain. This stuff will bounce off the plastic, giving almost no results, apart from heat build up. Grit with sharp edges will act like a razorblade shower, chopping the surface of the plastic. Thermoforming plastics usually have good impact resistance, so sandblasting them with round beads is similar to hitting a car tire with a hammer. Boing boing. On some plastics this may lead to surface flaking.
Abrasive diameter. Determines the amount of punch per particle. The bigger the particles, the more kinetic power, the bigger the mark they leave on the plastic, but it lessens the chance of getting them embedded in it. Sandblasting embeds your abrasive in the plastic. While it's usually non-reactive and carries no health risk (aside from breathing it in during sandblasting in large quantities), it may dislodge during use and get into the components of your device. Corundum is 9 on Mohs hardness scale. This may decrease the lifetime of your switches if it gets inside of them.
Try various pressures. More isn't always better. Less pressure means less abrasion, but also less heat, and less embedding.
Low temperatures make thermoplastics more brittle. This is actually good for the purpose of sandblasting, you might want to try that. Untested, though.
Don't let the abrasive get into your lungs. I think that's obvious, but it's a health hazard. Concentration causes permanent damage, leads to pneumoconiosis.
If you hope to remove the stripes from 3d printing with sandblasting, bad news. Not gonna happen, not easily anyway. You might want to pre-sand manually, or use some sort of a filler, to smoothen the print out first. The easiest way is to learn to stop worrying and love the stripes.
That covers the basics. I know there are quite a few engineers around, so if I made any mistakes, please correct me.
Now it's time to repost! The full thread is here, as of 2016-11-16: http://www.mig-welding.co.uk/forum/threads/diy-sandblasting-blasting-plastics.22944/
Oldie but goldie.
And below the extract of the most important stuff, with source attributions, + own commentary. I'm reposting this solely for the purpose of not having it lost in the abyss of the internets. If I goofed up, moderate plz, or tell me to remove. This info is pretty hard to find, and hence rather precious.
J-Blast Supafine (iron silicate/copper slag) can tear up ABS, can give raggy texture (rtbcomp, Nov 19, 2010 )
(that's actually super nice for reactive paint!)
[own commentary:]To avoid embedding, there were recommendations for dry ice blasting and soda blasting. Dry ice evaporates, soda is solluble in water/decomposes above 50degC, changing particle diameter (products are CO2, H2O and alkaline Na2CO3). IMVHO both are too soft, though low temperature may work in favor of dry ice. If you're adventurous, you can try kitchen salt (rinse it out), or ammonium carbonate (heat to 60 deg. Celsius to remove, it turns into NH3 + CO2 + H2O. Do it outside.). If you go through these, post results in thread!
Don't use ordinary sand. Just don't. Poor quality, contaminated... generally bad news.
"Blasting the clear acrylic with aluminium oxide seems to be giving quite a nice frosted finish" (kat.rawlings, Dec 10, 2010)
Quote from kat.rawlings, Jan 5, 2011, summary of their research, minor edits. Permalink: http://www.mig-welding.co.uk/forum/threads/diy-sandblasting-blasting-plastics.22944/page-2#post-259667
[materials tested were] Nylon 66, White PVC, Clear PVC, PE, PC and PP[...]
- Saftigrit (Aluminium Oxide) 180/220. Good results but discoloured the plastic so tested this in white and also in courser grades.
- Honite 13 Glass Bead. Not hard enough to significantly modify the surface.
- Blasting Grit. Good results and machine marks etc removed. The surfaces of the samples don’t feel as smooth and soft as when the aluminium oxide 180/220 was used. Again, the samples were discoloured.
- Kiln Dried Sand. Results not very good with a rough surface finish and there are also safety issues concerning kiln dried sand and free silica so it is probably best to avoid this abrasive.
- White Aluminium Oxide 180/220. Good results with a smooth, even finish, less discolouration than with the brown aluminium oxide but the surfaces of the samples appear glittery. This was the best abrasive for Nylon 66, polycarbonate, cast acrylic, clear PVC and white PVC.
- Plastic Media [from Hodge Clemco] (plastic and grade unknown[...]). Good result on polypropylene but the other materials feel rough to touch. This was the best abrasive for polypropylene.
- Garnet. This abrasive is far too coarse, leaving a deep profile in acrylic. It also discoloured acrylic.
- Glass Bead 801 AC. Didn’t have much effect on acrylic even when the psi was increased.
- Glass Bead AB 177-297 (courser grade than the one above). Still not hard enough to achieve the desired surface finish of acrylic.
- Brown Aluminium Oxide 80/100 (2 grades courser than the Saftigrit 180/220). Results were good but other abrasives were better on all plastics.
- White Aluminium Oxide 70. Too course- surface feels rough so the finer grade is better (180/220).
- White Aluminium Oxide 120. Again, the surface feels rough so the finer grade is better (180/220). However, this abrasive gave the best surface finish on polyethylene.
- Type 2 Urea Formaldehyde (Guyblast) Size 20/30. This blast media didn’t really affect the surface of the acrylic or polypropylene samples so it appears as though something coarser is needed.
- Type 5 Acrylic (Avialite) 30/40. This abrasive barely affected acrylic. The surface of the polypropylene sample was improved but machine marks were still visible and the samples don’t feel very smooth.
- Type 5 Acrylic (Avialite) 20/30. The 20/30 acrylic gave very similar results to the 30/40 acrylic media on polypropylene samples. It therefore appears as though plastic media will not give good enough results on polypropylene.
Aaand that's it for now! Good luck making stuff :)
-
Lot of good information. Thanks for posting it.
-
This answered all of my questions about sandblasting. Thank you for the 3D printed stuff.
-
I've been sandblasting more than a thousand keycaps by now, shared some of my experiences here:
https://deskthority.net/workshop-f7/keycap-sand-blasting-t12946.html (https://deskthority.net/workshop-f7/keycap-sand-blasting-t12946.html)
Found a great material (glass beads) and a good diameter to give ABS and PBT caps a nice surface. This will not restore the original GMK surface but remove any shine and sticky feeling of the keycap.
-
Bookmarking this thread for the future.:)
-
Just stumbled upon this thread trying to find more info on sandblasting acrylic.
I was wondering what nozzle size and nozzle pressure you were using in your testing?
-
@mhd,
Acrylic = PMMA?
From what I see it's routinely blasted to get a frosted glass effect. I didn't sandblast it myself, though.
Customary pressure for sandblasting is 4-6 bars. You can regulate abrasion with distance to the workpiece.
Nozzle size is usually determined by your compressor. The bigger the nozzle, the higher the air consumption, the harder for the compressor to keep up. I run at 6 millimeters, but most of my work is blasting off paint and rust off metal.
More important than the nozzle size is the diameter of your grit beads. I'd stay below 0.3mm max, but that really depends on the effect you wanna get.
You could try sodium as the blasting medium, it's transparent, and you can remove it with water. Process is called sodium blasting.
Corundum has a really expensive awesome transparent variant.
Aluminum oxide and pumice are white.
Test it. It's the only way to be sure :)
You might want to hit up wodan, he probably has more experience with plastics than me.