Splotchy anodize or dye result

[sccc]

This is my first post to this forum, would appreciate some advice as to my problem. This was my first (test) attempt at anodizing, and things seemed to be going well, but the final result was less than optimal.

After dying, the surface had some splotchy/blotchy spots, particularly around the perimeter of the piece. They are faint, but noticeable in reflected light. These are actually the BETTER looking part of the finish. I didn't notice if these spots were visible on the piece before dyeing.

Pictures and details at:

http://www.hogheaven.com/hobby/plating/anodize/firsttry.html

What's the most likely cause of this?

[neilfj]

The most obvious thing at this point is the anodization temp. It is 10-15F too high. Ideal range for LCD is 70-75F. Higher anodizing temps promote faster dissolution, which could explain your results.

Also, judging from the photos, you made the connection to the aluminum pieces by using sometype of square piece of aluminum. The shape can be seen in some of the photos. Those photos that have color within this square indicate a poor connection to the aluminum. The area under this 'square' anodized and blocked the current from reaching the remainder of the piece.
You have to make sure that the connection to the piece being anodized is watertight.

It may be an illusion, but the smudges in the center appear to resemble finger and hand prints, while the darker edges seem to be what I've run into when I've allowed the part to dry while it is still being processed. This is especially true when removing the part from the dye tank. The metal is at 140F which causes the surface to dry out quickly. It also occurs during other steps, so what I did was between each step in the process, keep the aluminum in a bucket of distilled water until you are ready to go to the next step. For instance, once you degrease with the SP, rinse, and let the part sit in a bucket of distilled water until you are ready to go to the de-ox step. Not only does this serve to keep the part wet, but adds to the rinse process and allows the temp. of the work piece to drop without drying it out.

[sccc]

neilfj, thanks for comments.

I realized that 80-85º was higher than optimal, but with daytime temps reaching 95º, that’s the temp that liquids in the garage seem to have equalized on. And the anodizing instructions say that higher temps result in more open pores giving better dye absorption, which is what I’m mainly after, so I figured it would be OK. I’ve got plenty of heaters, but no chiller. Perhaps I’ll toss some food out of the fridge next time, and chill the electrolyte off overnight.

Regarding the connection, I believe it was good. Having lurked here for a while, I’ve seen all the posts about importance of good connections. Connection was made using an aluminum electrical breaker panel connector tightly screwed to the piece. Continuity both before and after the anodize was equal to or less than the resolution of my DMM (0.01 ohm). The piece was not fully immersed, but held up so the connection was above the electrolyte. Some electrolyte creep did occur, as seen by the dye along the edge of the connector, but never appeared to get between the connector and the piece. The square attachment area is finely delineated by the dye, and that area still has 100% continuity, so I believe it was good.

Voltage during the anodize time would seem to confirm this. When power was first turned on, voltage started at @ 1.4v, within a minute or so ramped up to @ 3v, and within 10 minutes had reached @ 6 v, then continued to gradually rise to the final 10v over the full duration. This seems to correspond to the curves in the LCD manual, and indeed the part did anodize. I was constantly having to stir the mixture to knock bubbles off both the piece and the cathodes. And interestingly, I could see the voltage drop a couple of tenths each time I did this, indicating that the bubbles were adding resistance.

The part took dye very well, so it did anodize, and overall I’m pleased that it worked this good for a first attempt.

The drying issue may be closest to the mark. I had not considered this before doing the test piece, and had noticed fast drying when I took the part out of the hot SP degrease solution. I tried to do the tank-tank moves and rinses as quickly as possible after this, not knowing what effect it might have. As you noted, the drying did occur around the periphery first. However, in my case, it never reached the main surface of the part, and I never noticed it spot up as in the photos.

I have to wonder about keeping it wet though. The spots and periphery areas are actually the BEST looking. They are bright and shiny with better color depth.

An interesting observation that I hadn’t mentioned earlier: the spots on the polished surface are horizontal to the piece, but the streaks on the blasted/brushed side are vertical, and run nearly the full length of the piece. This is difficult to see in the un-retouched photos, so I’ll do a few contrast enhancements to highlight it.

I’ll do another test in a couple of days, and keep the temp w/in optimal range, and keep the part thoroughly wetted.

Thanks again. Sean

[neilfj]

I haven't seen the instructions you mentioned, but higher temps are definately not good. Higher temps increase the action of the sulfuric acid, which means that the anodized coating dissolves faster. This dissolution increases the pore size of the coating, as you stated, but it is not good that the pores get too large. It dissolves the cell from the inside of the cell, weakening the walls and opening the pore so large that the dye will not stay put. It seeps out and the sealing cannot completely seal the top of the cell. It is one of the 'problems' with anodizing. Cells to small and the dye isn't absorbed...cells to large and the dye seeps out. Its a balancing act, and good dyeing is only accomplished between the two extremes.

There are a couple ways to handle the heat in this weather. Take a larger bucket and put the anodizing tank in it, surround it with water and lots of ice. You only have to get the temp down about 10-15 degrees, so it should take too long. Or, as you stated, put in the refrigerator for a while. In either case, you'll have to keep an eye on the temp, as lower temps can also cause problems (cells too small to accept the dye). I've run into dyeing problems at 62F, so you only have about a 10F range. 70-72F seems to be ideal.

You know the importance of the connection, but from the photos, it appears that in some of the pics, that some of the area under the squares took the dye, which would be indicative of a connection problem. Although, I may be misinterpreting the photos. Yes, the anodization does migrate above the level of the electrolyte. I've found it migrates about 1/2"-3/8" above the electrolyte.

I didn't have time to look closely at your numbers the 1st time. You anodized .11 sq/ft at .500 amps (equates to 4.5 amps/sqft.). I'd raise the current to .66 amps. This will give you a current density of 6amps/sqft. I've gotten slightly better appearance with 6amp density than with 4.5.

Your anodizing time should be fine. Anything between 60-90 min should give you 1/2 - 3/4 mils based on the 720 rule.

Even with all this, since you don't know what alloy aluminum you are using, you may still have problems.

[Fibergeek]

Sccc,
Very good work for your first attempt.

I agree with Neilfj in all that he said; but let me say it again, you need to get the electrolyte temperature down to the correct levels.

Increased temperature also increases the conductivity (has less resistance) of the electrolyte, the reverse is also true. Your anodization voltages look a bit low due to the elevated temperature. You should expect them to be 1-2 Volts higher when the temperature is right.

A couple of other things regarding the spots:

Unless its been recently machined flat; flat surfaces on aluminum will have depressions and bumps all over the surface, some are too small to see until dyed. These surface defects will effect the uniformity of the polishing, the high spots get more polishing then the low spots. The low spots also tend to trap polishing compound, which can be surprisingly difficult to remove. ANY trace of polishing compound is trouble.



If your DMM will really go down to 0.01 Ohm, the test leads will add 0.1 Ohm unless you have a way to zero it, or unless you are using a 4 wire Kelvin connection. In any event, if you can measure less than 1 Ohm (test leads not zeroed) or less than 0.1 Ohm (test leads zeroed) your continuity test for the anode connection is valid.

[u[/u]

[sccc]

neilfj, thanks again.

Regarding temperature, I've been working from Caswells plating instructions. For the earlier, non-LCD, method they list a range of 65-80º, w/72º being optimal, so I fudged a bit and figured 82º would still give reasonable results. In the LCD specific instructions, they only mention the 70-72º optimal temp., but I presumed the range would be appropriate to both methods. Is the temp that much more critical w/LCD?

I had also assumed that one could manipulate one or more the variables to compensate for others that were not w/in range. e.g., if the temp were higher than desired (softer, open pores), you could run at a higher current (harder, closer pores), and the 2 results effectively cancel each another.

I've also realized I have a natural chiller. Over night temps still drop to 55º here in the northwest, so I'll just set the bucket outside one night, and run the test in the early morning after (the obvious is always the last thing you think about!).

Regarding the connection, I think you've misinterpreted the photos, and I could have explained it better. If you look at photo 4, right side, the 1cm.sq. area centered on the hole is the only connection spot. As you can see, this area is clearly delineated by the dye. Electrolyte did creep up the sides, but the connection point itself shows no anodize growth, still has 100% continuity, (w/in 0.1 ohm, see my next response to fibergeek), and absorbed no dye whatsoever. The aluminum connector itself shows similar characteristics. The exposed surface did get some wetting by electrolyte, and did take some dye, but the mating surface is clean w/ full continuity. So I have to presume the connection was good. I'll try something different next time.

I used the 4.5 amp/sq.ft. because it is recommend as the best compromise between hardness and dye absorption. Caswells recommended range is 3-6 amps, with the lower giving softer, open pores, w/more dye absorption, while the higher current gives the harder surface at the expense of dye absorption. I might have to try a higher current density as you suggest.

re: the unknown alloy: isn't that always the case w/hobbyists? The "real life" parts that I plan to anodize are of unknown alloy, and since the manufacturer has been defunct for 40-50 years, there's no way to find out. I just have to wing it here, and try to get an acceptable test finish and see how it works, then try the real piece. The "real" piece(s) are of cast alloy tho, so I need to find something similar to test with first.

Thanks for your help.

[sccc]

fibergeek, thanks for the comments!

As I’ve replied to neilfj, I will better control the parameters next time and post new pictures of the results.

The test pieces I’m using are cut from an old 1/8” plate which I’ve had for a number of years, not recently machined at all. After polishing it with the “wadding” polish, I had sprayed & rubbed it off w/simple green, rinsed it and left it to soak in hot SP degreaser for 15 minutes. Since it passed the water break test, I presumed it was good to go. But perhaps not? Will SP degreaser not get at embedded polishing compounds? What's better to use?

Notably, the reverse side was only brushed and blasted, so there was no polishing compound/chemical there, yet it shows the streaked finish. A different effect than the polished side though.

Whatever I’ve done wrong will hopefully be resolved w/tighter temp & rinse control next time.

Regarding my DMM resolution, you are right. I have my decimal point off by one. My bad, I should have checked the numbers more carefully before posting. The resolution is 0.1 ohm, and I use that AFTER taking test lead resistance out of the equation.

With test leads clipped together, the meter reads steady 0.9 ohm. With the leads about 1 cm. apart, w/in the connection area, the meter fluctuates between 0.9 and 1.0 ohm. With leads about 3 cm. apart, on either side of the connector area, the meter fluctuates between 1.0 & 1.1 ohm. So I had presumed this is the natural resistance of the aluminum. Is it possible that even this low a resistance is interfering w/anodize current flow?

With one test lead on the anodized area, continuity goes open circuit, so I know I’ve got SOME degree of growth there. I don’t have any means of actually measuring thickness, but I’ll try a crude scratch test later, and see how hard it is to break through.

As a trivia after thought, I wonder how the “big boys” do their anodizing to get 100% coverage. Until I started to learn about anodizing, I never knew that it was an insulator. But I’ve got a few pieces in my scrap pile that are 100% covered. These are some @ 1/16” plates, 3x10”, some dyed red, some black, which I picked up from an aerospace surplus place, just for possible project use. They appear to have been from a production test, since they are stamped w/alloy identification, e.g.. 7075-T6 & 6061-T6. I just took a closer look at them, and there is nowhere that I can find any spot where a mechanical or electrical connection was made. Even the edges appear 100% anodized. Is this possible? How’d they do that?

Thanks for your help!

[Fibergeek]

I would suggest using some hardware store aluminum barstock, say 0.25" x 1" or similar. This is practically guaranteed to be 6061 or 6063, the most common alloys in use, and has excellent anodizing characteristics.

Like you said, anodize is an insulator. This means that it is physically impossible to have "100%" coverage. You aren't looking close enough. The pros have gotten real good at hiding the connection point(s).

Connection resistance obeys Ohm's Law exactly, 1 ohm of connection resistance will drop 1 Volt for each Amp applied, exactly.

You should add some old fashioned vigorous scrubbing to your cleaning process, you'll find that it works wonders. Perhaps M_D or Easttex can make some specific recommendations, since these guys (among others here) are anodizing for commercial purposes.

[M_D]

..

[M_D]

The temperature very possibly was a factor, as the other guys have said. We have got some very nice results at 80-82º, but it is on the edge and if other parameters are off much failure is likely. The light spots on your part may because the electrolyte (acid) began to damage the areas anodized surface, it rarely would be even over the entire surface. Have you tried to wipe the streaks and spots off? Sometimes you get residues on the finished part that wipe right off, or at least with a little effort come off. It is important to rinse thoroughly and especially when using hot solutions you need to get the parts out and into a rinse tank in the least amount of time possible.

In our operation, the parts are pre-cleaned as needed to remove all visible contamination, and then cleaned in a final hot solution. Hand scrubbing with brushes and detergent such as Dawn dish soap in water as hot as is comfortable to work in, with rubber gloves, is effective also.

Getting the parts clean, in other words free of all oil, grease wax and etc., so water will sheet, might be only half of the required preparation, depending whether there is native aluminum oxide present. Aluminum oxidizes readily so unless the oxide has recently been removed somehow there is going to be some amount present, and it is next to impossible to tell by looking at the parts how much or how thick the oxide is. The oxide retards the formation of a good anodized coating, and if the oxide varies in thickness, you will get an uneven color, with the light areas showing where the heavier native oxide was. Parts that have an even oxide coating may turn out with an even appearance, but it is harder to get the colors deep. We do a lot of fresh and fully machined parts (the entire surface is machined), in which case we do not have to de-oxidize them, and they come out looking bright and shiny. So in addition to cleaning the parts of oil and etc., for best results the oxide has to be remove so all that is left is clean bare aluminum.

The oxide can be removed in various ways. The Caswell stripper, or a mixture of sodium hydroxide and water, will remove the oxide, and etch the part to some degree. You have to experiment to get the right mix ratio to suit as it affects the final look. A weaker mixture (say a few teaspoons to ¼ cup per gallon) takes longer, 10-20 minutes depending on temperature and will leave a flat surface (low gloss), and a strong (12-16 ounces lye per gallon, use with caution) will tend to do the job in 10-60 seconds and leave a brighter surface. But once it starts to bubble hard, if left in longer it will dull and flatten the surface also. If the part gets a light film of smut on it from the lye, it should be sufficiently de-oxidized. The smut can be easily removed with the Caswell desmut solution.

As far as 100% surface coverage it is true that it can’t be done. Most commercial anodizing is done with titanium racks. Depending on the parts size and what it require to fully secure it on the rack, a couple of small sharp contacts may be all it takes, but there will be at least at least a couple of little pinpoints that aren’t anodized.

Fibergeek, this is a photo of some of the parts I used the CD Welder on that I told you about, I have higher resolution copies if they would do you any good.

[Fibergeek]

Thanks for the cleaning advise M_D.

Could you post the low resolution picture (conserving Caswell's BW) in the Sput Welder thread? Please add also any observations and/or comments you may have on it.

[sccc]

Gentlemen:

Sorry my reply is delayed, but other priorities intervened. Thanks to everyone, for all your advice.

I've run another test (photos here), and as usual, Murphy was there. The finish came out blemish free this time, but now the color is off. Two steps forward, one back!

I changed my procedures to comply with everyones recommendations. The piece was kept thoroughly wet at all times. Electrolyte temp was 72º at the start, and had risen to 75-76º at the end. Voltage ramped up as expected, rather quickly at first, continually slowing over time. It peaked at 11.2v this time, at about 70-72 minutes, and began slowly dropping. At 75 minutes voltage was down to 11.0v, so I figured PAR had been reached, and pulled the part out.

As you can see from the photos, the blemish problem is gone, but the dye didn't take nearly as well as the first attempt. So I'm guessing the temperature difference caused this. Since I prefer the deeper color of the first test, next time I'll drop the current to try and get the deeper color back.

M_D, what a great looking piece of fabrication! What color blue did you use on those parts? That’s just about the color I’m chasing after.

Thanks everyone!

[M_D]

sccc, it sounds like you made some progress. Also, thanks for the compliment on the parts in my photo. I believe that blue is called Electric Blue. We can get anything from a light tint to a very deep, dark, and rich blue with it, somewhat like the darker shaded areas look like in the photo.

Let me give you a few hints on color depth. First off, when you change one thing, like the current density, you usually need to at least change the anodize time, unless you were just out of the ballpark beforehand and need to bring one factor in line. If you lower the amperage, you would need to increase the time to keep it relatively even with what you just did. To make it darker, you may or may not need to increase the anodize (or dye time if it was short before) time even more. There is a saying that you can't change anything, without changing everything. That is, every change will affect the whole deal. You might even try anodizing at the same current, and increase the time another 10 minutes, or increase the current and anodize for the same time.

Different acid concentrations in the electrolyte, electrolyte temperature, and current density affect the pore structure. Generally, to an extent, the thicker anodized coating gives a deeper color potential. There is a maximum point though; I think fibergeek reported it is like 6 or 7 tenths thick. If you have the proper balance of acid concentration, temperature, and current density, you can juggle the anodize times and extend them further than you have done, without running into trouble. Some pore structures take and hold the dye better, or simply give a different look. For example, you can put a thick coating on and dye it a deep color, with the results looking a lot like a paint or powder coating. With a few little changes and a thinner anodize coat, you can get the same deep color, but with more of a metallic look.

I know it might sound over whelming, but the bottom line is there is quite a bit of latitude when anodizing, and experimentation and practice will help you to learn the boundaries and effects of certain parameters. Once you find a sort of happy medium where the parts turn out relatively consistent, then you can start adjusting the parameters and seeing where they take you. One thing I will say is this, consistent preparation and control of the variables is a must. Otherwise, you may have either an exceptionally good or bad result, and blame or credit the wrong thing, which can lead you astray. Once you get lost, you need a baseline to go back to. I think that is one of the best things about the LCD formula, if it is followed faithfully you will get good results. If not, there was something out of tolerance to find and adjust.

[Fibergeek]

Wernick, Pinner, and Sheasby in The Surface Treatment and Finishing of Aluminum and its Alloys report on testing for dye penetration in an anodized layer. The bottom line is that dye (any dye) won't penetrate any further than 0.7 mils (0.0007") regardless of what you do. For dyeing purposes, there is no point in an anodize layer thicker than 0.7 mils.

[Sid03]

related to part of this post....

Caswells desmut will also de-oxidize, correct?