Very High Grit Sandpaper Application Question

[bostonman]

Recently I was shopping for spare sanding pads for my palm sander and saw the variety of grit numbers.

Last year I built a bookcase and used 220 (or maybe 400) between coats along with a final sanding. That sandpaper was very smooth and provided a very glass like feel on the wood.

I'm curious, when the grit number gets up to 1000 or higher, does it actually do much? Seems like a liquid polish would be more suited rather than the possibility of a hard particle getting between the sanding pad and scratching something.

[Kleinstein]

For wood the much finer grids than 220 are rarely used, mainly for oiled surfaces of hard wood up to maybe 400.
Besides the grid also the glue used can make a difference to the feeling and finish. Epoxy glued is usually sharper, more grip, but also trends to cause deeper scratces than natural thermoplasic glue.

The realy fine ones are a thing for the final touch before high gloss paint/varnish and for polishing metall or plastics.
The liquid polish can be easier to produce and avoid scratches. The finest "paper" (actually plastic film) I remember using was 6000 grid, but this is alerady rare and paste may be more common then.

[Bud]

Working on minor car finish scratches need 2000-4000 grit before polishing with paste. Same for knife blades, but may need to start with lower grit, depending on blade condition.

[bostonman]

I remember using was 6000 grid

Wow. I thought 800 or 1000 was smooth.

Working on minor car finish scratches need 2000-4000 grit before polishing with paste.

Good point. Anytime I hear "sandpaper" I think it's going to scratch (in this case) paint on a car whereas a liquid polish is the necessary solution.

Someone at an old job built a hobby radio and made the knobs on a lathe. If I remember correctly, he used fine grit sandpaper, wet I think, and then a liquid polish. It was smooth and looked like chrome.

That was the first time I ever heard of very fine grit a high grit number sandpaper; but I don't remember the number he stated.

[Bud]

1000 grit will eat car clear coat in no time. Better to start with 2000 and then go up. Fairly deep scratches will take you forever if only to use polishing paste. Also, using sandpaper on some sort of a supporting block (wood or plastic or foam one) gives beter control over the job. Polishing paste is messy, you lose visibility of where you go, and typically it is used with electric polishing tools and fairly large polish pads (2-6 inch) so it goes over large areas. Polishing paste is used as final touch.

[Analog Kid]

I often use 1500 grit to polish metal and plastic.
It's way too fine for wood, though it could be used for a wood finish like varnish or lacquer (not oil).
My sequence is usually something like this:

o 150 (for gross flattening)
o 220
o 320
o 400
o 600
o 1500
o rubbing compound
o polishing compound

Everything above 320 is the black wet-or-dry silicon carbide paper.
o Flitz

[Psi]

Sandpaper stops feeling like sandpaper and just feeling smooth around 1500.

[coppercone2]

I did a test piece of wood(small table) to 5000 grit and it ends up looking funny like its made out of plastic (with Poly)

It looks better if you just do what carpenters on youtube do lol

Weird matte but really smooth finish. I think it looked really cheap.

[shapirus]

The denser (and harder) the surface of the material to be sanded, the higher grit of the abrasive can typically be used with it to still be able to achieve an increasingly finer surface.

For example, to achieve a mirror surface on a steel part (e.g. handmade fine craft knives etc.), and I mean actual mirror finish, not just glossy/shiny, meaning that it's properly flat or otherwise uniform, of a very low roughness and without any scratches visible to the naked eye at any angle, one would go progressively up in grit to at least P2000, maybe beyond, and only then use abrasive pastes (of course with a respective technique, not just rubbing it against a "polishing" wheel on a bench grinder, but that's another story).

Wood, of course, being soft and porous, will not typically benefit from higher grit polishing, unless maybe when it's real hard, like that exotic stuff that's illegal to export, and/or covered with a particular coating that can benefit from it.

Another example: when natural sharpening stones are prepared, they are first lapped (made flat), and then "conditioned" by rubbing them in a certain fashion on a flat surface (typically a sheet of glass, granite or cast iron) with some loose abrasive powder to set a specific roughness to their working surface. It is not uncommon to use SiC powder as fine as F1200 (same as ANSI 1500) on a final stage, which is roghly equivalent to about P6000-P7000 sandpaper grit (~3 micron particle size). Similar to sanding, it typically only makes sense to do this with hard types of stones, one of the most commonly known examples being translucent Arkansas.

p.s. yes there's a shitload of various grit standards. One could go insane just trying to convert one to another, and they often aren't directly comparable, too.

[Analog Kid]

I did a test piece of wood(small table) to 5000 grit and it ends up looking funny like its made out of plastic (with Poly)

It looks better if you just do what carpenters on youtube do lol

Weird matte but really smooth finish. I think it looked really cheap.

Depends on the wood.
Hard, dense tropical hardwoods like rosewood, ebony, ipe (goddamn, that stuff's like hardened steel!) can be polished to a high gloss with fine sandpaper and polishing compounds. Other softer woods, not so much.

BTW, was the wood you tested finished or raw, with no finish (varnish, lacquer, shellac, etc.)?

[bostonman]

Sandpaper stops feeling like sandpaper and just feeling smooth around 1500.

That's what I wondered. It seems it's almost a sheet of paper at around 800, so I wondered how effective it actually is, but makes sense it's for polishing metal.

I hand sanded my bookcase I think with 400 grit and that came out smooth like glass (poplar wood), but I've never tried sanding metal to a chrome finish.

[Analog Kid]

BTW, tangent discussion here, but I do some (amateur) woodworking and wood finishing.

I don't like using sandpaper at all if I can avoid it.

Turns out that the smooooothest finishes on wood are those that come from a sharp plane or a cabinet scraper.
The tool cuts the fibers instead of shredding them as sandpaper does.

Gots to be razor-sharp for that to work.

[Analog Kid]

I hand sanded my bookcase I think with 400 grit and that came out smooth like glass (poplar wood), but I've never tried sanding metal to a chrome finish.

You should try it sometime; it's actually very satisfying to produce a mirror finish on a piece of metal.

Steel, aluminum, copper, brass will all finish to a very high gloss.

It's not necessarily easy. Takes a lot of time and patience.

Think of the process this way: each grade of abrasive (sandpaper or polishing compound) leaves scratches in the material you're polishing, but the scratches get progressively finer and finer.

The trick is to make sure all the previous, larger scratches are completely gone before going to the next finer grit. If they're not, you need to go back with coarser abrasive and get rid of them.

The last thing I use to achieve a mirror finish on metal is this great stuff called Flitz. Available here in the US, the abrasive is actually made in Germany, so I'm sure either it or something very much like it is available elsewhere in the world. It's a bluish cream that gives an absolutely brilliant mirror finish after using fine polishing compound (like the typical white automotive polish or rouge).

[bostonman]

What about "wet" sanding? Is that just taking a piece of sandpaper and wetting it to reduce friction? i.e. less scratching

[Analog Kid]

What about "wet" sanding? Is that just taking a piece of sandpaper and wetting it to reduce friction? i.e. less scratching

Yes, if you use the "wet or dry" paper (typically, or always in my experience, silicon carbide, the gray or black stuff).

The water floats the sanding swarf away so it doesn't clog the paper and form blobs that create larger scratches. Has nothing really to do with friction (sanding is basically all about friction).

You can use soap, but I've always found just plain water to work fine.

[Psi]

Sandpaper stops feeling like sandpaper and just feeling smooth around 1500.

That's what I wondered. It seems it's almost a sheet of paper at around 800

Na, even at 1000 it still feels 'grippy', it feels like it will remove/scratch material.
A sheet of paper is probably more like 2000-3000

[jpanhalt]

Places that sell automotive refinishing products (paint stores) will have the very fine grits.  I always use them with water.  3M has a line of products that are graded in microns and fractions thereof and have a plastic backing.  Years ago, I used to them to finish polish glassy carbon electrodes for an electrochemical HPLC detector.

[CatalinaWOW]

Two things drive grit size.  The basic structure of the underlying material.   And the way you are sensing the surface. 

Wood is a composite material with a sort of periodic structure of harder and softer material (the grain).  You can't really make it much smoother than the size set by this structure which is why woodworkers don't really use the higher grit numbers.  Here you are often using your fingers as the sensing method.  While you can sense discontinuities (steps) of a couple of thousandths of an inch (0.05 mm) you can't really feel small areas at this scale.

When polishing mirrors, glass or plastic for optical use you need to get the surface smooth to roughly a wavelength of light.  Or about 0.02 thousandths (~.5 micrometer).  Shiny paint finishes similar.  Hence the much finer grits.  (By the way this information is useful when reclaiming scratched CDs and DVDs.  Most can be recovered by polishing the scratches out.

As stated earlier you start coarse and work your way down.  You don't have to start coarse, but if you have any large or deep scratches you will need a calendar to measure you progress towards a completely smooth surface.

Polishing creams are not well graded but are equivalent to grit numbers in the thousands, typically multiple thousands.  So aren't usually a good place to start.

[Poroit]

G'day bostonman,

Have a look at articles relating to French Polishing for Grit size as that gives a very smooth finish.

https://www.instructables.com/How-to-French-Polish-1/

The trick to a real smooth finish is to keep raising the grain and sanding repeatedly.

We used Methylated Spirits then set it alight.

[bostonman]

This all makes more sense.

My question started because I used several 80 grit pads for my palm sander on a project and went to Harbor Freight to buy more. In the past I saw high grit pads, but this time I paid more attention because they had a larger variety in stock.

While there, I thought it may be worth buying a few different grit numbers for potential future needs, and felt the 800 (maybe they had higher grit numbers too - not sure now) because they had smaller quantity packages with a small section of the box open to allow feeling the pad.

It didn't seem practical to buy anything other than the variety I have (60, 80, 120, and 220); the 400 grit (?) are sheets I used to sand my bookcase by hand.

When sanding metal with >1000 grit, is it all done by hand? In the case of the person who made knobs for a hobby radio, I'd assume the knob was being spun on something (a drill?), but this seems as if the loss of control due to spinning would result in scratches, however, spinning by hand wouldn't give even full turns causing it to lose its round shape.

[shapirus]

When sanding metal with >1000 grit, is it all done by hand? In the case of the person who made knobs for a hobby radio, I'd assume the knob was being spun on something (a drill?), but this seems as if the loss of control due to spinning would result in scratches, however, spinning by hand wouldn't give even full turns causing it to lose its round shape.

It depends. When you need to polish a flat surface (flat as in no shine-through light will be seen between it and a flatness gauge), then yes, there's basically no way to do it other than by hand, unless maybe you have access to some specialized (expensive) equipment.

Otherwise (especially with shapes created by rotation), power tools can be used of course, as long as you know what you're doing and have the proper tools and technique. In any case it is necessary (for best results) to maintain a uniform pressure and equal amount of removed material in all places.

[Analog Kid]

When sanding metal with >1000 grit, is it all done by hand?

Yes and no. Depends on what's being polished.

Small items, yes, probably best to do by hand.
Larger items, like say a large flat metal sheet, might be better done with a power tool. However, this doesn't mean attacking it aggressively with a polishing pad in an electric drill and expecting good results.

Some industrial polishing processes involve moving a rotating polishing wheel, charged with compound, back and forth across the surface in a regular pattern for consistent results. You can replicate that by carefully moving a spinning polishing disk back and forth in a regular pattern.

Another method is to use a polishing cloth wrapped around a flat block and moving it back and forth.

Most small metal objects can be polished very well just by rubbing it with a cloth.

There are also specialized processes you can try where a small polishing disk is pressed against the metal and then moved a regular amount to give a circular patterned polished surface; you could rig up a drill press to do this, with some kind of fixed stop system to move the workpiece a constant distance each time.

[bostonman]

This gives a new perspective to any possible projects involving metal and needing shine along with being smooth.

I double checked (not that it matters), but the sandpaper I have is 400 grit and what was used to give a final sanding to my bookcase. If I remember correctly, 220 between coats and 400 as a final (although I my have used 400 too between coats).

This was the highest grit I've used, so hopefully someday I'll tackle a project that allows for >1000. Most likely if it's a project that important, I may not use Harbor Freight sandpaper.

When NASA made the lens and additional lens for the Hubble Telescope, I wonder if they bought the sandpaper at Harbor Freight or Home Depot.

[CatalinaWOW]

Even though they initially did the Hubble wrong, I doubt if sandpaper of any grade was ever used in the manufacture of those lenses and mirrors.  Optics is a very broad topic with many different processes and machines used.

I have used a great deal of sandpaper from dirt cheap and up in my life, and haven't found that the grit size was meaningfully wrong in any of it.    What does vary is the life of the sheet, how quickly it plugs up and how long it cuts well.  Those things make the better grades well worth the money.

That said, you can get a headlight polishing kit from Harbor Freight which I have found works perfectly well for removing the frosty haze that develops on older car headlight lenses.  It includes an arbor, several grades of sanding disc starting at 400 and going up and also some final polishing compound.  Works on CDs and DVDs also.  An inexpensive start on finer finishes.

[Analog Kid]

I'm planning on polishing my headlights, but I won't be buying any kit from HF or anywhere else, as I already have all the materials I need: 320/400/600/1500 wet-or-dry silicon carbide paper, rubbing compound and polishing compound. Oh, and nice soft clean rags.

[jpanhalt]

320 for headlights?  I would start at 1500.  If that doesn't work get coarser until it does.  You will need to finish higher than 1500 unless you have power polishing.  There's a difference between shine and scratch free optics.  For example, electropolishing can give a high shine on a relatively rough surface.  When you get to the end, a power buffer gently applied will be appreciated.

[Psi]

I'm planning on polishing my headlights, but I won't be buying any kit from HF or anywhere else, as I already have all the materials I need: 320/400/600/1500 wet-or-dry silicon carbide paper, rubbing compound and polishing compound. Oh, and nice soft clean rags.

as jpanhalt says, 320 is too course you will add scratches only to then have to remove them.
Perhaps if your headlights are deeply scratched you might need 600 or 800. but not 320.

Do a test on a deep scratch to figure out how fine you can start with while still removing the scratch

[coppercone2]

the problem is sealant they always go bad after a while

supposedly it has some anti oxidant coating. I tried the kits with the liquid but it still goes hazy after a while. if your lights are at this point, it turns into a maintence procedure. Not a bad way to hold off another year or two, but their degraded

[Psi]

the problem is sealant they always go bad after a while

supposedly it has some anti oxidant coating. I tried the kits with the liquid but it still goes hazy after a while. if your lights are at this point, it turns into a maintence procedure. Not a bad way to hold off another year or two, but their degraded

It's almost like they were better 40 years ago when they were actually made of glass.

[jpanhalt]

Almost?  Glass was better, but maybe a few pennies more to have molded in today's shapes.   My 1965 VW had glass lights that you could replace without reprogramming anything.  Despite all its so called conveniences, I do not like my 2024 Mazda CX-30.  There are a few things that stay set between starts, but some others are complete nagware.  Cruise is a real annoyance.

[Analog Kid]

320 for headlights?  I would start at 1500.  If that doesn't work get coarser until it does.  You will need to finish higher than 1500 unless you have power polishing.  There's a difference between shine and scratch free optics.  For example, electropolishing can give a high shine on a relatively rough surface.  When you get to the end, a power buffer gently applied will be appreciated.

You're right, of course.
I just put 320 as a starting point because it's the lowest grit on my scale; I would probably start at 600 for my headlights. I have patience, so don't mind spending more time on each grit to remove previous scratches.

I'll look for grits above 1500 at my local home improvement store (not Home Despot!). And the polishing compounds start off from there (coarser "rubbing compound" (reddish brown), "polishing compound" (white)).

[Monkeh]

3M do a very simple kit which is 500, 800, and 3000 grit, plus a polishing pad and compound. Just add drill, takes about ten minutes a light. Easier than digging up all the loose bits..

[jpanhalt]

Your home improvement store probably won't have high grits.  Any auto supply that sells finishing materials will or should.  Aside from industry, their main use is "color sanding" auto finishes.

[CatalinaWOW]

If you have the patience to start at 1500 either good on you or you are very picky about your headlights.  The problem isn't scratches, it is a thick layer of degraded material.  Can't say. Whether it is UV exposure, oxidation, outgassing or whatever.  On my twenty year old vehicle it took quite a while with a power tool and 400 grit to cut through the yellowed crud.  Then you can start polishing which actually goes pretty quickly. 

I agree that it becomes a maintenance chore, but between the lack of availability for older lenses and their high cost it is well worth it.

[mzzj]

I remember using was 6000 grid

Wow. I thought 800 or 1000 was smooth.

if you need really fine grit the magic word is "lapping film" instead of "sandpaper"

3M lapping films are available to 0.05 microns that translates to about 500 000 grit. This would be wayyy beyond what is considered mirror polished.

[shapirus]

3M lapping films are available to 0.05 microns that translates to about 500 000 grit. This would be wayyy beyond what is considered mirror polished.

Wow. I wonder what the application would be. UV optics?

[jpanhalt]

As mentioned, I used them to make electrodes from glassy carbon for an HPLC detector.  They worked very well.  I believe the smoothness may have decreased absorption in the detector.  The main advantage by far was that glassy carbon allowed virtually any solvent to be used as compared to graphite paste that was used before.  I did not go to 0.05 micron as I recall, but the grit I used was finer than DuPont color sand paper.

[Kleinstein]

Very fine sanding / polishing is wanted for metalograpic smaples. This can be optical, AFM and SEM. Usually the fine part it done with pastes, but a paper can be faster.

[mikerj]

320 for headlights?  I would start at 1500.  If that doesn't work get coarser until it does.  You will need to finish higher than 1500 unless you have power polishing.  There's a difference between shine and scratch free optics.  For example, electropolishing can give a high shine on a relatively rough surface.  When you get to the end, a power buffer gently applied will be appreciated.

It depends on if you have to remove the remnants of the original anti-scratch coating or not.  It can craze badly and get patchy on old headlamps, and you'll be there for days trying to remove it with fine grit.  The finish doesn't last long if you have to do this though, even with a UV protector, as polycarbonate is so soft.

[jpanhalt]

That's precisely why you start with fine and get coarser, at least for someone who doesn't do the job everyday.  Once it's hit with coarse grit, you are committed.

As for crazing, my only experience is with acrylic windshields.  In that instance, it was due to a combination of stress, wrong sealant (mineral oil-based is bad), cleaning with organic solvent based cleaners, and exposure to sunlight.  It was not not a coating and was not removed appreciably by polishing.  A picture at the outset would have helped.

[Analog Kid]

The finish doesn't last long if you have to do this though, even with a UV protector, as polycarbonate is so soft.

I wonder if this wouldn't indicate hitting the headlight plastic with some kind of clear coat (like lacquer) after polishing.
Would have to be done very carefully, of course, to prevent runs, sags, orange peel, crazing, etc. Not to mention masking off surrounding stuff.

[jpanhalt]

I know you said "like," but I would never use solvent lacquer on plastic.  A water-based clear coating like epoxy or preferrably polyurethane would be safer (e.g., Minwax polyacrylic).  The problem is crazing or fogging.  It might also be an odd use for acrylic floor wax.  That wouldn't be very durable, but it is easily removed and replaced.

[coppercone2]

glass is slightly less translucent then acrylic, so you have attenuation. Maybe it was a tiny benefit when the lights were still low, but the LED are so bright now, who gives a crap, they say its blinding drivers.

What would be nice is sapphire crystal like a rolex. Maybe this is the improvement we will see with cheap energy that makes it cheap to make high grade ceramics and glasses in cars of the year 2500. But I think they used to be able to say that the light is measurably brighter when acrylic housings were new, for marketing. That feature turned into a hindrance now.

The only issue might be thermal management, its possible if glass does attenuate more, it might get too hot, its possible its a issue with brighter lights. But you would think if this is true, then the scratched up fogged lights would be melting

https://longislandwatch.com/sapphire-crystal-mods/

I guess a headlight would be like $2000

https://www.bmw.com/en/magazine/design/swarovski-crystal-headlights-in-BMW-i7.html





ah but it seems that most lights are made from polycarbonate, which is less translucent then glass, not acrylic.

[CatalinaWOW]

It is the stylists and industrial engineers, not the opticians and physicists who are setting the material specs for headlights.  Well there is some tiny contribution from the aerodynamicists. 

Somehow the car buying public has been convinced that the shape and visual appearance of headlights is really important.  So just about every model has a unique shape.  Hence relatively short production runs (tens to hundreds of thousands, not the millions to tens of millions for the old sealed beam headlights).  And complex shapes which are harder to mold in glass.  Hence polycarbonate and perhaps other plastic materials.

Super bright LED light sources don't really help, because the problem with older headlights isn't optical transmission, but scattering.  While the transmission might have gone down by 10-20%, the light delivered 50-100 meters down the roadway has gone down by a factor, perhaps in the neighborhood of ten in a severe case.  The rest of the light is off illuminating the clouds and the fields and buildings off to the side of the road.

[bostonman]

Super bright LED light sources don't really help

Not sure if this is country wide, but in MA, seems the new thing is driving with the high beams on all the time at night (they don't even turn them off when you flick your high beams at them), so doesn't matter about super bright LEDs or hazed headlight lens.

Then one car has super bright white headlights, another has blue, etc...  Seems the standards for headlights have changed. Some SUVs have headlights so bright you'd swear they have their high beams on; maybe they are aftermarket but have heard from owners their headlights are stock but people flick their high beams at them thinking they have their high breams on.

The other new way of driving, at least in MA: sit at a green light (most likely because the driver is staring at their phone or 52" control monitor screen on their dashboard), but go through a red light that has been clearly red for a solid three-seconds.

[SteveThackery]

Another use I have for lapping films is for tuning fountain pen nibs. I use 5, 3, 1 and 0.3 micron films. Obviously the initial shaping is done with 5 micron, then progressively finer to polish the nib until it gives the desired amount of "tooth" (resistance as it moves over the paper). For a toothy nib I usually stop after using the 5 and 3 micron films. Applying the 1 micron film makes the nib noticeably smoother, and the 0.3 micron film lets the nib feel like it's almost floating across the paper.

Although a floaty nib would seem desirable, many people (including me) find that their handwriting is neater when the nib has some drag on the paper.

PS: toothiness is not the same as scratchiness. A scratchy nib tears the paper fibres and drags them along with the nib, causing blotches and other unwanted marks.

[jpanhalt]

Do you  polish differently for lefties and righties to correct for the angle and strokes?

[SteveThackery]

Do you  polish differently for lefties and righties to correct for the angle and strokes?

The experts - "nibmeisters" - do, yes. I don't because I'm just a hobbyist doing my own pens.

[jpanhalt]

Esterbrook didn't (circa 1950's).  Was it not an expert?  I'm glad you know the difference.  I would return to ink in an instant if I could find a decent pen at a reasonable price for a leftie.  Otherwise, I am stuck with Pilot ballpoints.