accuracy of the 720 rule

[kebo]

I have logged dozens of anodizing runs using different alloys, different current densities and different target oxide thicknesses. For each run, I calculated the run time using

Time(minutes) = CD/(thickness * 720)

After each process I measure the oxide using a calibrated isoscope, and have found that the oxide is typically +-7% of the target oxide thickness. So my question(s) is/are:

1. Just how accuate and reliable is the 720 rule?
2. What are some factors that will affect the accuracy? bath temperature? conductivity? Rectifier ripple?

thanks
kevin

[RedRiver]

The 720 Rule is: 720=ASF x minutes / mils. Where; ASF is Amps per square foot, mils is the desired coating thickness (1 mil = 0.001"). Therefore to solve for the duration in minutes the formula must be rearranged into the following : Minutes = 720 * mils. / ASF

Example : Solve for duration at 12 ASF and a 1 mil thickness.
This is how I would solve for the duration using my formula.
Minutes = 720 * mils / ASF
Minutes = 720 * 1 / 12
Minutes = 720 / 12
Minutes = 60


Do you notice the difference between my formula and your own?
Minutes = ASF/(mils * 720)
Minutes = 12 / (1 * 720)
Minutes = 12 / 720
Minutes = 0.01666

Now...Assuming that you made a type-o when presenting your forumla I will try to explain what may cause thinner thicknesses. (As to how I understand it).

One should consider ALL of the factors. The 720 rule is the formula used to calculate the anodizing duration and/or the projected outcome and it assumes the following : that your electrolyte is maintaining a constant temperature of 68F, that the electrolyte's concentration is exactly 9.4% H2SO4 by volume, that your surface area is calculated to the exact square inch and that you are utilizing an anode:cathode ratio of 3:1. Other factor such as alloy may have an effect on your results as well. Without meeting the exact specifications you will find some variation in your results. If what you suggest is true and that you are only achieving 7% of the expected thickness I would hazard to guess that your acid concentration is too high or your CD too low. It would seem that your anodic layer is being dissolved almost at the same rate as it is grown. Either that or you did not make a type-o when presenting your formula and you are anodizing for far too short of a duration.

Just out of curiousity, plug some numbers into your formula to give us an idea of what you're working with...It may help.

I hope this helps clear up the mystery somewhat...
- Dan

[sswee]

I read +-7% of the target oxide thickness as the target thickness +-7%. Which if you have a target thickness of 1 mil or .001", +-7% would only be .07 mil or .00007". If this is the typical variation on the 720 rule calculations, I consider it to be extremely accurate. In the machine shop, parts with a designated tolerance call out in the tenths of thousandths also have a temperature call out for the part to be checked at.
As for the parameters that effect coating thickness, Redriver covered it well.
SS

[RedRiver]

Originally Posted by kebo

After each process I measure the oxide using a calibrated isoscope, and have found that the oxide is typically +-7% of the target oxide thickness.

Perhaps I missunderstood what kebo meant in the above statement. If I understand you correctly sswee, you understand kebo's statement to mean that he is within +-7% of his target thickness? I understood kebo's statement to mean that he has only achieve +-7% of the projected thickness.

Having said that, if you are within +-7% of the projected thickness kebo then job well done! I would think that any of use would be very happy with those results. On the other hand, if you have only produced an oxide thickness of +-7% of the projected thickness, then perhaps one of the factors mentioned above is a little wacky.

I am somewhat unclear as to what you are asking kebo...Then again, it doesn't take much to confuse us canucks! Just show us something shiny which makes a "whirling" sound and we'll tilt our heads in utter awe and confusion.

[kebo]

thanks for the feedback,

Dan,
I did indeed have a typo in the 720 (need to be more careful in the future).

In all of my process runs, I record the number of amphours consumed and I know the target process area very precisely (most of the runs are done using 12" x12" coupons of 7075 or 6061)

To find my thickness discrepancy i rearrange the 720 as follows

thickness = (CD x time)/720
= (Amps x time)/(area x 720)
= (Amphours x 60)/(area x 720)

one run I did yesterday consumed 46.0 amphours on 2 sqft.

thickness = (46.0 x 60)/(2 x 720) = 1.92 mils

using the isoscope however, I measured 2.26 mils for a discrepancy of about 18% (18% is the largest discrepency I have see, and as I have stated, most are within 7%)

I alway thought temperature, concentration, anode/cathod ratio, and alloy would change things, and so now I wonder.....

If all of the variables were held constant, I should get consistant result and be able to compensate accordingly.....

and

what is acceptable in industry? If 1 mil is required what is the acceptable tollerance? As sswee stated, +-7% maybe very accuate, but is it accurate enough? (trick question that can only be answered by the person that asks)
thanks
kevin

[RedRiver]

In regards to your question whether +-7% projected thickness is accurate enough I would hazard to guess that the answer for that question would lie with the customer. I would think that a customer who supplies an engineering drawing for their part would have a noted tolerance on the drawing. In my books, +-7% would be just fine!

Just my $0.02
- Dan

[kebo]

thanks for the feedback....I will consider the issue closed
kevin

[2scoops]

Originally Posted by RedRiver

(1 mil = 0.001").
- Dan

Why is 0.001" called a mil?
I was getting a bit confused then thinking of millimeters. I always call 0.001" a thou and 1mm a 'mil''.