Quick question on "old" ano method......

[Fibergeek]

Sid,

Since you are using CC the electrolyte concentration won't make much difference in the thickness of the anodize. This is not the case with voltage (CV) anodizing. The concentration will effect pore size. I would prefer a 1:2 ratio at 12 ASF, and still use 30-45 min. If everything else is right, this will get you 0.5 to 0.7 mils coating thickness. Recall that dye won't penetrate any deeper than 0.7 mils. regardless of pore size.

The anodizing time is from the "720 rule". I have found; since I can measure the anodize thickness, that it holds up well at any ASF from 3 to 24 ASF.

The measured voltages look OK for 1:1 electrolyte, you should expect to see higher voltages (a little higher) with weaker electrolyte.

As M_D and other Ti racking users will agree here; racking in general is tricky, the two main issues:

1. Connection variance; your spring tension or how and where it makes contact may not be right, or not right most of the time. The sporadic nature of the problem supports this.

2. The parasitic resistance of the rack itself. If the SA of the rack itself is close to or larger than the work, it dominates. Since the Ti won't anodize here this resistance is constant with anodizing time; the resistance won't go up like it does for the work. This can defeat CC; the ASF for the work actually going down, not remaining constant, the power supply can't tell the difference between the rack's resistance, and the work's resistance. A fudge factor added to the ASF will help, I don't know how much extra ASF you will need to compensate.

Just as a test to see if this is the problem; do a test anodization with a piece of 6061, with soft Al wire bolted tightly. I know you can't do this with your real work, this is only to isolate the problem.

By "hazy film that you can sometimes wipe off"; could this be described as "chalky"? If so; it would be a sign of excessive dissolution, and the stronger electrolyte is working against you. Other things that can do this are; too low ASF, too high a temp., too long an anodization time.

[Sid03]

You may be right about the spring tension of the Ti racking, but at this point I wouldnt say thats it. The Ti racking I have seems to put a fair about of pressure on the parts.....more so then what I previously achieved with "rigging" fasteners out of the soft 12 gauge Al ground wire I used. Ill run another batch today at the 1:2 ratio and see how things go. Something tells me that the 2011 alloy has a part to play in this also, since 6000 series Al almost always turns out better!

[Fibergeek]

2xxx series has a very different anodization curve than 6xxx series, the high Cu content of 2xxx doesn't help either. You should avoid anodizing different alloys together, this will complicate things. I'm not sure why this is, but LCD tolerates the traditionally "hard to anodize" 2xxx and 7xxx series better than conventional anodizing.

Try the experiment I suggest above, if we can isolate the problem we can fix it quicker and more effectively. It will be worth the effort.

[Sid03]

could lack of cathode area also give results like a described?

I noticed that the cathode that I directly connect my PS lead to puts off alot more bubbles then the rest. I had the others connected together with some fairly decent test leads.....I think the wire thickness may mostly be insulator now that I look closer at them. I ran some 12gauge Al wire to all the plates and not all the plates seem to be putting off about the same bubbles.

If the cathode plates were only "working" very little before, how would that have effected everything?

[M_D]

We are anodizing 2011 on Ti racking too, and it’s definitely more troublesome for us than 6061 and 6063. At first it gave us fits, now we are quite consistently getting nice results and things seem to go better and better. The commercial anodizer that we have used was especially inconsistent with it (2011), they didn’t like it either. I will say that over 90% of the 2011 we do is dyed black, the balance is fairly evenly divided between clear, red, and blue. The red with 2011 is by far the hardest to get right.

I realized sometime back that the Ti racking makes it difficult if not impossible to calculate the exact amperage that the parts are anodizing at. We started out using 4.5 amps with aluminum hanging wire. Once that was going fairly well, we needed to increase volume and went to Ti racking. We still don’t have a large operation (shopping for a 1,000 amp power supply now), and the most parts per run are 100-200 parts. We started to anodize at “6-8” amps per square foot (1:2.5 electrolyte), to reduce cycle times. The reason I put quotes on the amperages is because that was the calculated amperage for part surface area. Anyway, we were doing ok (not super great) after some experimenting on times and so forth. Then we went to a deeper and larger anodizing tank, and were able to fully submerge our racking (we were only submerging it about 40-50% before that). All of a sudden things got weird. I decided to try more amperage and we are now using 10-12 amps per part square foot (not including racks). It made a huge improvement, the parts are turning out very nice and have very little or no haze that needs wiping off. If we have problems with haze now, it is almost always with 2011. The electrolyte is now the same (1:2.5 electrolyte) as we used with the smaller tank.

It appears to me that anodizing with Ti racking does require compensation in the amperage amount. Too low of amperage causes haze (smut or what ever you call it) and other inconsistencies like black dye being reddish, and etc. Since the Ti racking is altering the electrical calculations, I’m not sure what the actual amperage for parts really is. Like Fibergeek suggested already, I would suggest trying more. I suspect that since you increased the electrolyte concentration so much in addition to the amperage, that you are still low on the actual part amperage and dissolution is occurring now and was before.

You do need to be sure the parts are de-oxidized well, because that positively will cause problems. You might try increasing the soak time, or concentration of the etch solution, if it doesn’t degrade the surface. Are you de-smutting them after the lye etch? I know that very thorough rinsing is needed, and racking multiple parts makes it more difficult than doing a single part that has free access to all sides.

I know none of this information is precise, but is might give you a direction to try. I think with the Ti racking you may need to stick with loads that are fairly consistent, and experiment with the amperage and times until you get good results. If you change the part load, then I believe there will need to be an adjustment on current density. Maybe with time Fibergeek and I will find a formula were it can be calculated before hand.

Added in Edit: I also thing that doing some test pieces with aluminum wire would help. If they turn out good and the results are consistent, and the Ti racked loads don't, it shows the problem is CD related.

I would think a deficiency in the cathodes would cause problems. I have been using aluminum sheet (stays in the tank). I have 2 pieces of 7 square feet each on opposite sides in a 55 gallon tank, connected with a 12 guage wire with solder terminals. The connections are several inches about the solution, but I try to clean them regularly as the acid mist finds it way around.

[Sid03]

M_D- thanks for the input.

I have been adding 1.5A to the current needed to try to compensate for the Ti racking. I may call the manufacturer tomorrow and see if I can get and exact surface area of the different clips im using. Do you guys think the exact surface area of the rack should be accounted for, or only a percentage of it since the Ti conducts differently?

I just ran a batch of 6061 and all looks well(with Ti racking). I'll try the 2011 again......orange and red, this should be fun!

[Fibergeek]

Nice and definitive, M_D.

To answer Sid's cathode question:
The possible cathode issue you raise would be a connection problem, not cathode area actually. If CC isn't being compromised too much by the rack SA, CC will raise the voltage to compensate for what it sees as lower conductivity in the anodizing tank circuit. This is why cathode area isn't critical at all in CC anodizing circuit. Yes it is a circuit and all the physics and electronics math does apply.

As far as calculating the exact compensation required by the racking it is certainly possible, using Ohm's Law and Kirchoff's Law of DC Circuits. You will need to know the SA of the submerged portion of the rack, the SA of the work, the conductivity of the Ti alloy, the change in bulk resistance of the work as it anodizes, and the conductivity of the electrolyte. To do it accurately the math will be messy, and the applied ASF will have to increase somewhat over the anodization time, not remain at one value. I think a simplification can be made, where only a fixed increase in current density will work well enough. M_D is probably close to that right now, but his increase in ASF would only apply to his setup and his racking. You will need to develop a fudge factor that works in your case.

[M_D]

[quote="Sid03"]M_D- thanks for the input.

I have been adding 1.5A to the current needed to try to compensate for the Ti racking. I may call the manufacturer tomorrow and see if I can get and exact surface area of the different clips im using. Do you guys think the exact surface area of the rack should be accounted for, or only a percentage of it since the Ti conducts differently? ………….quote]

I wish I could answer that with certainty, but I can only say that I don’t think you want to compensate for the Ti with the amps per square foot as the aluminum. We have been using the 720 rule for some time now. When I try different current densities, I use this factor to estimate the time needed to comparably match the thickness of parts done at another density. Right now, we are usually going to 600, which seems to be working well (parts are dying nice, and no real problems are evident). So if we are using 10 amps per square foot of parts, we anodize for 60 minutes (60 x 10 = 600). If we use 12, we go for 50 minutes (50 x 12 = 600), and so forth. On the 6061, for red we go a little longer to a full 720, and it seems to give a noticeable difference in the depth of color possible when dying red. It may not be the thickness per se that is helping; it may be the pore structure instead. I don’t know for sure why, it just helps.

One difference I notice between 6061 and 2011 is this; the 6061 seems to be more tolerant of “over” anodizing than the 2011. One thing that is puzzling though, is that 2011 doesn’t build the thickness as fast as 6xxx series. But if we anodize the 2011 at the same rate and time as 6061, it looks like a good match. There again, it may be the pore structure balances it out, rather than actual thickness. But, if we anodize longer than usual, the 6061 takes longer before it deteriorates than the 2011 does.

On some of our parts, the racks even exceed the total part area. If the Ti needed as much current as the aluminum, I think that we would be needing to add more current or add more time, if you compare the dying tendencies we see with Ti racking against parts anodized with aluminum hanging wire where the wire area is either insignificant, or calculated in and compensated for.

At some point, I will carefully do some identical parts with both aluminum and Ti racking, and see what I learn.

I know this is just general and highly unscientific, but it is working. The unfortunate thing about it is not being able to factor it all like the LCD instructions outline. If one were doing odd batches, where the part surface area to rack surface area varied constantly, then it might be pretty tough to maintain any predictability at all. As it is, we are using either a little different current or time for various parts, to compensate for the area ration between the racks and parts. It has been from trial and error, seeing the tendencies, and adjusting accordingly. That’s why Fibergeek has rightly referred to it as Kentucky windage.

[Sid03]

Well, my batch of 2011 turned out ok. I kept in for about the same time as yesterday(50 minutes) but changed the ano bath from 1:1 ratio to 1:2 ratio, and ran a better connection to all the cathode plates. Once again, the voltage decreased while anodizing. It started at 17V and ended at 11V. The bath temp also went up about 3-4 degrees, dont know how much that effected it.

Do you think the fume control balls actually help maintain bath temp much(I read that they do somewhere)?

How bad is the hydrogen gas to breathe? Its getting very noticeable again since im not using LCD anymore. If I put a papertowel inside a regualr dust mask I can work fine without noticing it, but without the mask your coughing. Im going to make some sort of fume hood, just wanted to know how dangerous it is until I get that complete.

Thanks guys

[Fibergeek]

OK. Glad to hear things are improving.

The voltage drop that you are seeing is characteristic of a 2xxx series anodization curve. 7xxx series does this too, but not as pronounced. 6xxx and 1xxx series have a curve that slopes up, and the coating forms much faster.
You don't want the bath temp. to get much higher than that, you should be thinking about some sort of temperature control.

That isn't hydrogen that you are breathing, hydrogen is ordorless, colorless and tasteless. The small amount of hydrogen generated escapes quickly (its the lightest element) and isn't any sort of hazard. You are breathing sulfuric acid fumes, the ventilation where you are anodizing isn't adequate. You will want to fix that fast.

Mist suppressant and balls will certainly help, but you do need good ventilation.