I have some hollow section steel beams that I want to paint. They are covered in black greasy substance that I think is black oxide and oil. I can degrease it, but do I need to remove the oxide? I've tried to google this, but all i find is the description of what this oxide is and its purpose.
It is "mill varnish" . Difficult to find out what exactly it is because different mills use different stuff and change from time to time; assuming it is relatively new steel. Big purchasers will tell the mill not to put it on if they are going to paint the steel, it is a nuisance to remove. Once you clean off the varnish you still have the mill scale ;black iron oxides that may be difficult to remove ( industrially it is usually done with a steel grit blaster). With exposure to weather it will start to turn brown as different iron oxides form( rust). If you have time ,one technique is to let it weather to take off black mill scale because the brown rust is easier to blast. An angle grinder is probably most practical to clean a few pieces in a non-factory situation.Without blasting , you should use a conversion coating , mostly phosphoric acid ( such as "Navel Jelly") .It does a good job of converting a little remaining oxide to something more stable , Then prime and topcoat.
Prep is a hard science
NASA operates the Kennedy Space Center at Cape Canaveral, Florida, also known as the “Cape who picked this location, Nikita Khrushchev?”
NASA being NASA, they’ve done staggering amounts of research on care of millions of tons of steel latticework in ocean-washed salt-marsh Florida. Many books written, and they even have a website dedicated to the subject: http://corrosion.ksc.nasa.gov
The short, short version is that the gold standard is grit-blasting down to “SSPC-PC10 near-white metal”, then following with a 2-part mil-spec primer, then a barrier coat of primer for an appropriate LPU topcoat. This is head-and-shoulders above every other option. A distant second is “aggressive grinding”, and everything else lags far behind that.
LPU is “2-part” paint which requires an A-part colored base, a B-part converter for brush, and a reducer for brush (you really do not want to spray that stuff), plus a matching primer (as it will hardly stick to anything, nor will anything stick to it). Pencil in $500 before you even get started.
However I have also seen 30-year performance from plain old non-lead alkyd primer and alkyd topcoats, when the prep was done to SSPC-PC10.
Conversely I also have stuff where I “winged” the prep, using hand wire brush, rust converter and the very best LPU primers and topcoats, and that stuff failed off within 5 years. Terrible.
The takeaway is, anything you do less than media blasting to SSPC-PC10 is a major step down.
This is steelwork. It needs steelwork paint.
Even if it’s architectural, you need to paint it like it’s a locomotive. So not latex paint.
I’ve been having a great deal of trouble procuring that mil-spec primer, but if I am able to blast to SSPC-PC10, I use a 2-part marine epoxy primer directly, followed by the LPU topcoat... and I’ve gotten away with it so far.
Otherwise I am resigned to premature failure of the primer. Grinding is too messy and firestarting for me, so my fallback position is to use heavy wire-brushes in an angle-grinder. It doesn’t remove all the rust but comes close. Then a 2-cloth wipedown with paint thinner, then - believe it or not - common Rustoleum 7769 Rusty Metal Primer. Applied on a hot, dry day with solar gain heating up the metal, so it is quite hot and virtually no water is being captured under the primer. I can’t explain why it works so well over medium prep, but it does.
Now, when I want a better job, I’ll overcoat it with the good LPU (and its own primer as a barrier coat; LPU won’t stick to Rustoleum).
However, I’ve also overpainted the Rustoleum with a plain alkyd topcoat, and that’s done acceptably well.
Like I say... Top-drawer prep with common paint is better than common prep with top-drawer paint.
One legal issue, though
Now, there’s a legal issue here. All these products are stinky as all getout, and have high VOCs approaching 720 g/l. Many Air Quality Management Districts restrict the VOC content of paint used for architectural use, i.e. for buildings. This is why latex paint has taken over (and also why it’s turned to crap, and is so sensitive to going moldy - the ~100g/l VOCs that used to be in latex paint kept mold from growing).
I don’t paint architecture much, and I haven’t gone through the gory details of all the regs. Generally they are looking for very low VOCs, however they do grant exceptions for a wide variety of cases - such as trim, and I would expect, steelwork. Some jurisdictions give you carte blanche for anything you buy in quart or smaller cans.
Find your area’s rules, and follow them.
The mill scale needs to go, yes.
The sticky varnish is to keep the steel from rusting enroute to you. If it’s hot-rolled it also has gray/black “mill scale”.
Mill scale is funny stuff. It actually does a pretty good job of rust prevention, and will work for about 40 years underneath primer and paint, but then, it will start becoming a rust vector. Any crack in the paint will allow in water, and that will cause the mill scale to convert into a bulkier type of oxide (that is also much easier to remove). That will let water advance under the paint, so you’ll get new peels and chips developing as the years pass. When it gets into this state, it’s incredibly annoying - you are repeatedly wire-brushing (to knock off peels) and repainting for appearance’ sake. The only way to stop it for good is to strip to bare metal.
Are you sure it is oxide or scale? Black If scale you want to remove it it is a byproduct of the mfg process we usually use a heavy wire cup brush or ~80 grit disks. I think the Twisted wire cup Cut the scale and last much longer than the paper but the cup’s cost about 12+ compared to 50c for paper? After grinding we wipe with a solvent , if it’s red that is a weldable primer it can stay.