Last Update: Jan 9, 2001
The pace and complexity of modern industrialized life preconditions us to believe in the need for expensive specialty and sub-specialty services for tasks that can be accomplished at home, with a little perseverance (although not on a commercially viable basis). As an example, take the restoration of old model IC engines. Many of these share a common problem; they have (or had) red anodized, screw on cooling fins to which a past person or persons unknown has taken a large set of multy-grips and a cold chisel, resulting in horrendous graunch marks. A good restoration would require the heads be cleaned up and re-anodized. After some experiments, my co-researcher and I achieved some remarkably successful results in color anodizing using the most primitive of equipment. So if you're curious, here's how. For collectors , the engines in question were two Australian Taipan 1.5cc (a round head and a flat top head), a Frog "Vibramatic" and two ME Herons.
0. Research! — in this case, magazine articles from SIC (Strictly Internal Combustion) and MEW (no, not Model Engineer's Workshop, the other British one, Model Engine World), plus the Tee Workshop Practice book. Naturally, all were somewhat contradictory and follow-up letters to the editor in MEW added more confusion. My description below (cut from an email to the "Motor Boys" fraternity) gives the process which is now working consistently for me.
1. De-anodize the old part by immersing it in a weak caustic soda solution—like a level tea spoon dissolved in an old coffee cup of warm water. Takes no more than 5 minutes and the part may turn "black" depending on the alloy. Keep this up until all, or virtually all the old color has disappeared and the outer surface is dull. Rinse under tap water and scrub with an old tooth brush to remove surface scale. Rubber gloves are probably a good idea for this (and other) process.
2. Mount on some kind of mandrel and skim it up with minimal metal removal on the outside (would you believe the Taipan head internal thread is 27 tpi? Some kind of pipe thread, I think. With only a QC box on the Myford, I cut a loose 28 tpi and that worked well enough). Finally, run at high speed with fine wet 'n dry using kerosene. I also buffed them on a Scotch-Brite belt that runs on one side of my bench grinder.
3. Bend a strip of aluminum to go inside the head/whatever, making good spring contact. This needs to be able to pass the anodizing current, so take a little care. The other end of the strip will be screw mounted to a common hanger bar connected to the positive side of the current source—hence it is the "anode" and we have the origin of the term anodizing. Aluminum welding rod is another hanger candidate, although I've not tried it.
Now back into a fresh caustic solution for about 30 seconds to de-grease, then rinse in cold, demineralized water. From the time it comes out of the caustic, avoid all direct contact with the part. The oils on your fingers will leave patterns on the part, as will the minerals present in some domestic water supplies.
4. The anodizing bath is 1:1 battery acid and demineralized water (my cloths dryer condenses water into a collector for me, so I've a never ending supply of this). The container should be clean and plastic. I used a 2 liter square food container with a snap-on lid for storage. I've no indication of how long a batch of electrolyte like this will last. So far, mine has been used ten times with no apparent change in the results, but anyway, we are not talking significant cost here—just be careful disposing of it.
The cathode used is a piece of lead sheet hammered flat from an old length of lead covered communications cable. It's bent in a U shape inside the bath, with another U bent so it sits on one rim of the container. The negative wire from the current source gets attached to it (soldered). The refs say it should surround the part (hence the U shape) and be at least of the same area. I went for massive over-kill in the cathode area and this has worked.
The aluminum hangers of part(s) to be anodized are then screwed to a bar (I used 1/8" al strip, but any thing would do) that sits over the top of the bath and carries the positive wire from the current source. Plastic cloths pegs either side of the bar prevent it accidentally touching the cathode. That would be bad.
The refs say about 65-70 degrees F for the bath. That's about room temp for water here in spring, so I didn't bother to heat in any way.
5. Now the techie-part. The anodizing current required is about 100 mA per square inch (all sources disagree, so this is a good, round number average). This apparently impacts the pore size formed and hence the ability of the oxide to accept and trap the color. You can laboriously calculate the surface area, or approximate it. From the heads I've done, a rule of thumb seems to be surface area = 2.5 times the area of the head calculated as a simple cylinder. I didn't bother to include the area of the hangers, but if they are big and flat, it would be a good idea to figure them in.
The current source needs to be DC and about 12 to 20 volts, capable of delivering at least 2 amps and variable in some way. You also need to be able to measure this. My flying buddy (who is the head engineer at a local radio station) "found" a supply that was close to this, but not variable, so we fitted an auto-transformer (Viriac) but a drill speed control to the mains side of it would also do the job. A battery charger driven by a speed controller would probably work ok. My current source has some filter capacitors on it which a battery charger wouldn't though.
After you have the aprox surface area for each part in the batch, add them up and divide by 10 to get the current (in amps) required. Hook up the source negative to the lead cathode; the positive to the black lead of your multimeter (which should be set to the 5 or 10 amp DC current range) and the meter red lead to the wire coming off the hanger with the parts to be anodized. Now plonk 'em in the bath and crank it up to about the reading calculated (err on the high side). Start the stop watch and observe the part. After a minute or so, a fine stream of bubbles should start to form.
Keep this up for 45 minutes to an hour, periodically checking the current and adjusting as required. At any time you can lift the bits out (turn off the current and hold via the bar). If they appear a dull gray, that's good. If not, run them a while longer. This is a real suck it and see operation. Avoid breathing when inspecting the bath—it's evil stuff (explosive, too). I did this out doors—I suspect the fumes would have attacked every piece of steel in the workshop.
When satisfied, rinse parts in demineralized water and place them out to air-dry thoroughly. This is important! I put them in the sun for an hour or so, but avoid all direct contact (oils in the skin etc). We are going to depend on capillary action to get the die into the microscopic pores formed by the anodizing, so the dryer the better. Even over-night would probably be ok. Believe me, going into the color bath wet gave no color take-up at all, even though one source advocated this!
6. The color bath is not a precise process. I had success with a product called "Dylon" used for dying fabrics that I obtained from the drug store (chemist shop). It comes in little round plastic tubs with a crimped on alloy cap and seems to be globally available. The color "Scarlet" works great—but their "Emerald" was a dismal failure, so I conclude that the dye particle size varies with the color. It it's too large, it won't go into the pores. I dissolved half a tab of die in a 1 pint jar of de-mineralized water to give a very concentrated mix. This is warmed in the micro wave oven for 1 minute before use (this step may not be necessary, but having a process that works, I don't want to vary it!).
Put the kettle on and start it heating to make steam, then pop the part in the dye (holding it only by the hanger) and dunk and swish it around. Try to avoid hitting the sides of the jar. At one stage, I bubbled compressed air into the jar, but swishing seems to work just as well. Take the part out periodically an look at the color absorption. When it seems like it's got all it's going to get, give it a bit more and make sure you've got steam. This will take only a few minutes.
7. Tap off the excess liquid, then place part in the steam coming from the kettle. Rotate, twist, turn, etc to get steam into all nooks and crannies. This is closing the pores, sealing in the color (and apparently changing the type of oxide formed from a hydride to a hydrate—at least I think that's what one of the refs said). After you've burnt yourself a couple of times and are totally sick of it, take part and kettle to the sink and pour the boiling water all over the part.
8. You can touch it now. Dry it off—I used paper towels—and expect a little color to rub off, but if you've got it right, not a lot. Finally rub some machine oil over it. This brings out the luster and changes it from merely nice to absolutely spectacular.
Thats it! As I said—the die particle size is the critical thing. My friend Peter Crewdson, from Canberra was staying with at the time (I'm RC, he's PC and *still* an aviation electronics tech with our FAA equivalent). He and I experimented with red and blue ink as the die as well as red and green Dylon. Only the "scarlet" Dylon worked acceptably, but it worked great. He's gone home determined to anodize the entire Canberra airport red, just for the fun of it.
If it fails, go back to step one and try again. Our test piece (a reject Schroeder Deezil head with a strange internal thread cut for no valid reason) went round this loop three times. The final anodizing is good, but the surface is dull due to fine pitting, probably from the multiple caustic dippings to de-anodize—so good idea to practice until you're confident and if you must de-anodize, re-polish each time.
I've got some blue Dylon to try, and the owner of the Frog, Herons and one of the Taipans who donated the power supply has just dropped an old Taifun on me with a tab of pink Dylon—he insists it had a pink head—but you'd never know to look at the sad old thing. That will probably be the next experiment, along with the stuff-up head for the Owen "Mate" (was busily turning out the id to match the of of the fin groove—got to within 10 thou of creating a set of rings before I realized what was happening!) which I'll try the blue on.
This is a real empirical process, but after we arrived at it, we did three totally successful batches in a row, over two days, so I think we have it down (for the setup we used, anyway). The more pure the alloy, the better the color absorption. We hung one batch on aircraft alclad hangers. The 99.9% pure coating came out an absolutely brilliant red, but on the edge you could see that the core (high manganese and silicon, I think) was much darker and duller.
Update 11/18/1998: I've been contacted by the author of the MEW article (Mr N Rat) who tells me he has recently restored a blue head AM 15 using Dylon "Kingfisher" blue with spectacular results. This seems to confirm my theory that the particle size in Dylon brand varies with the color as my blue experience was not so good.
I've also tried varying the current/time ratio to see how this effects pore size and hence, color absorption. I think I can definitely say using a higher current (up to 130 mA per square inch) gives much better results than lower (less than 100 mA). This conclusion was derived from results observed in red anodizing three batches, so they're not exactly conclusive, but good enough that I'm now going to use higher current for a shorter duration as standard practice.