Fibergeek's Anodization Curves, 1 of 3

[Fibergeek]

This is a set of curves for different alloys at 18 ASF. I will explain them as well.

http://www.caswellplating.com/bbs/album_pic.php?pic_id=234

This is 6061, 2024, 7075, and 1100 (as a reference). For anodizing purposes, they will be essentially the same for any 6xxx, 2xxx, 7xxx, and 1xxx alloy.

The current density was held to 18.000 ASF (yes, 3 decimal places) in all cases, and the electrolyte temp. was held to 70.4 deg. F. +/- 0.3 deg. F. (my cooling system will do this).

The electrolyte was the "professional" 9.4% by volume mix.

Notice that the curves slope upward for 1100 and 6061, the "easy" alloys. The curves for 2024 and 7075 slope downward, the "difficult" alloys. Also notice that each alloy has its own characteristic bulk resistance (OSF).

The vertical axis is in Ohms per sq.ft. (OSF). You can calculate what the voltage would be for your size of work as follows:

OSF x 1/SF = Ohms, this simplifies to: OSF / SF = Ohms

The voltage for your work is then: V = A x Ohms

SF is the surface area of the work in square feet.
A is the actual current you apply to the work to get the ASF that you want.

What makes the curves slope up or down is the combination of coating thickness and pore size, more thickness makes it go up, and larger pore size makes it go down. Unfortunately we can't separate them and look at each individually while the anodization is proceeding.

The curves that slope up show that coating growth is staying ahead of dissolution. Notice that these have the thickest coatings. When the curves slope down dissolution is winning the race, and the coatings are thinner.

This means for 9.4% (by volume) electrolyte, better results could be had by increasing the current density (speeds up growth) reducing the acid concentration (reduces dissolution) or by decreasing the temperature (reduces dissolution). Increasing the current density would be the easiest to do.

BTW, the 2024 and 7075 samples both showed signs of excessive dissolution (powder coming off when dry) 2024 was the worst.

The graphs are in the Album (button at the top) how do you make this work?

(edited to include the formulas)

[Sid03]

since the 2024 never goes uphill, that means that dissolution is always winning out? How does a layer form at all under these circumstances, since according to the graphs dissolution is always occuring the whole way through the process?

also curious as to your "6061 vs. ASF" graph. Im guessing the thickness measurements were taken at the end of 90 minutes? Wouldnt the 12ASF have been thicker around 40 minutes since dissolution started to occur? Should the ASF's about 12 also seen dissolution above 60 minutes or so?

[Fibergeek]

No.

2024 forms a layer 17 um (0.67 mils) thick in 90 minutes under these conditions. The only way I can think of to monitor coating growth alone is to remove and measure the sample at say 5-10 minute intervals. Or to have a number of identical smaller samples, and remove and measure them one at a time at regular intervals. The applied current will have to also be reduced incrementally so that the current density remains the same.

Most of us have experience with the evil effects of excessive dissolution. Ironically, if you were to reduce dissolution to zero you would not form the columnar structure that makes anodization useful. Some dissolution is absolutely necessary, the trick is to not have too much.

You need to understand what this "Ohms per Square Foot" (OSF) stuff means, if you just ignore it you will never be able to correctly adjust the anodizing parameters to suit your needs. You will be stumbling around in the dark, by random chance you might stumble on some particular combination that works OK. You won't know why, and you will be back to square 1 if you change anything (like the size of the work) just like a "professional" anodizer.

Anodize is aluminum oxide, a ceramic thus an excellent insulator. If its an insulator, then how do you get current to flow through it so that the anodization will form?

The resistance observed (more correctly bulk resistance) is the conductance of the electrolyte when it is restricted by the hollow columnar shape of the pores. All pores that can make any electrical contact with the base metal and the electrolyte contribute to this. Its magnitude is affected by pore size, coating thickness, and surface area. When the pores are larger, more electrolyte can make contact, thus the resistance looks smaller. When the coating is thicker, the electrolyte column in the pores is longer, and the resistance looks larger. When the surface area is larger, there are more pores involved, so the resistance looks smaller.

The correct way to interpret this set of curves is to recognize that the alloy content (particularly the copper) in 2xxx and 7xxx series results in slower anodization growth. This allows the dissolution more time to open up the pores before the desired coating thickness is achieved.

OK. How to apply it.
You want to adjust things so that the down sloping curves slope up, or at least flat line. It will never be as good as 6xxx or 1xxx, but substantial improvements can be made.
First thing; 9.4% electrolyte is obviously too strong for 2xxx and 7xxx, its fine for 6xxx and 1xxx, if you reduce the acid strength, you reduce the dissolution rate.
Second thing; lowering the electrolyte temperature will also reduce the dissolution rate.
Third thing; raising the current density will make the coating form faster, it doesn't effect dissolution rate.

If some combination of these three things is done the anodization of 2xxx and 7xxx alloys can be improved, probably a lot. These changes will of course effect other things, we will deal with that in the remaining two sets of curves.

[Sid03]

interesting. For awhile I was running 6061 and 2024 in the same batch.....I thought the 2024 ano'd before the 6061. Reason being, the 2024 had signs of dissolution while the 6061 looked fine.


given your experimentation so far, do you think im in the ballpark as for running both 2000 series and 6000 series in the same setup? Bath is 1:2 ratio. 12ASF. Usually run it around 40 minutes. I am using Ti racking.....just a small rack. I usually add an extra amp to my current to help compensate for the rack. Everything seems to be pretty decent......I havnt had any noticeable dissolution problems. The 2000 series is taking dye very well, and the 6000 series does as well. Im hoping im close to being happy with my setup. The only thing im worried about is how thick of a layer im actually growing........I have no way that I know of to check it other then taking a nail to it.

[Fibergeek]

Let's withhold any judgement on how you are doing things now. You will need to see the rest of this before you can make an informed decision.

An eddy current type electronic coating thickness gauge is indispensable for any sort of serious anodizing. There are also types that use a magnetic field, these work on steel only, don't get the wrong kind. I got mine on ebay about a year ago, about $425.00 used. You can get new ones now that are better than mine and cost less than $425.00 new.

Sid,
I'm going to suck you into this project; what I need from you is the make and part numbers of the titanium racking parts that you are using. I'll also need to know the number of these parts in your particular rack, and about how much of it is in the electrolyte. With this we can get a reasonable idea of the rack surface area in question.

What you get when we are done is a definitive answer; how much is your racking effecting your setup, and how much extra current (or whatever) do you need to compensate for it. We'll be using your setup as the numerical example. We will be able to validate the results by coating thickness measurements on pieces you anodize. I'll measure them and return them to you.

M_D and anyone else out there using Ti racking are also invited to participate, I'll need the same racking information from you guys.

To everyone else:

My anodization setup is for research purposes and uses electronics lab instrumentation. This degree of control and accuracy is necessary to get good data. For any sort of practical anodizing (non-research) my setup is 10-50 times more accurate than is required. Don't let the numbers scare you. You can do entirely suitable anodizing with your Chinese CC/CV power supplies.

[Sid03]

totally willing to participate. Is there another way to contact you to send some of this info? I tried to PM you and apparently it is disabled. Thanks

[M_D]

It sounds like a worthwhile mission to me. Besides knowing the rack surface that is submerged, what is else do we need to document beside the standard parmaters of current density, tank temp, and electrolyte concentration?

[Fibergeek]

Sid,
I'll contact you via email, I'll get that through Caswell.

M_D
I don't see a need for anything fancy or really any different in the data you take for normal anodizing operations. What will be different and very important for this to succeed, is getting a good idea of the racking surface area thats in the electrolyte. That's why I'm asking for make, model, and quantity of the racking elements in your "sample rack". This should be pretty much what you use regularly, unless you wish to assemble a special unit.

The other thing needed is the surface areas of the parts you regularly anodize, the alloy, and the number of them in a typical batch. Again it can be what you normally process anyway to minimize the disruption.

We will need coating thickness measurements and I'll do that. Sid and you will need to send me a few pieces from the sample batch (2 or 3) and 1 un-anodized piece for calibration. I will return them to you.

The piece of Ti rack and the Ti hardware that M_D sent me a while back will be used to establish some basic electrical conductivity measurements for titanium.

(edited to add)

I forgot to mention, M_D's setup will be the second numerical example, like Sid, he will get numbers that apply to his setup specifically. They will be different than Sid's.

[Sid03]

my batches contain varied parts, none of which are the same surface area. If anybody has a cheap place to pick up some similiar sized and series aluminum blanks or extruded tubing, please let me know. Ill look around and see if I can come up with some useful parts for this.

[Fibergeek]

The parts don't have to all be the same. All that is necessary is to know the surface area of each and the alloy. Some 2011 would be good since its generally the most troublesome. 6061 (or 6063) should be the other alloy, because its so common.

I believe that both Sid and M_D are using both of these alloys.

Sid, I have an email address for you at earthlink. I'll send you a test email this afternoon.

If no one objects, I'll arrange for the three of us to trade contact information, is this OK?

[M_D]

Fibergeek, I will need your shipping and e-mail address, so if you still have mine could you send it to my e-mail? I lost some hard drive data recently due to a power outage or power surge.

I do have 2011, 6061, and 6063 parts. I will probably run parts for this myself so I know exactly how they were done, rather than send parts that were done by employees that I can't trace to any certain data.

Also, I will figure some surface area of racking. On some of our racks that are laminated in several layers, it may be difficult to establish with absolute precision since there are gaps between some portions of the laminations. But I should be able to get close.