Scratch Resistance/ Hardness

[Falstaff]

We have one of our AR coating machines which occasionally fails a scratch resistance test. The Bayer Ratio from lenses coated by it are always a little lower than than that from our other plants. We can find no physical problems with the ion source, and have tried adjusting various parameters. However, nothing that we have done so far seems to make any difference. Is there anyone out there that can suggest something else that we could try?
The process is normal physical vapour deposition (not ion assisted). We have neutralisation current comparable with that from the other plants. The test lenses are lacquered indentically to those going through the other machines. Is there anything other than the ion source that could be to blame? There are no alarms. The ion source is a Lion mark 1, and is generally regarded as small for the size of the coating chamber (Boxer).

[chip anderson]

Don't know a lot about this subject other that what I have picked up from experience. But some things scratch easier because they are harder an inflexible. Automotive paint for instance is made to be intentionally flexible to prevent scratching and chipping. Could we be looking for too much of a good thing in scratch cotes?

Chip

[scriptfiller]

I plead ignorance in the technicalities of the A/R coating process, so I will hypothesize..

Are the enviromental conditions identical plant-to-plant, internal temperature and humidity? Are the raw materials being sourced from the same vendor from the same process lines and the same warehouse? Are the machines being ran at the same electrical systems, voltage, amperage and wattage making one machine run "hotter/colder" than the other?

Are the substrates being "lacquered" in the same manner?

Some shots in the dark.

[Barry Santini]

We have one of our AR coating machines which occasionally fails a scratch resistance test. The Bayer Ratio from lenses coated by it are always a little lower than than that from our other plants. We can find no physical problems with the ion source, and have tried adjusting various parameters. However, nothing that we have done so far seems to make any difference. Is there anyone out there that can suggest something else that we could try?
The process is normal physical vapour deposition (not ion assisted). We have neutralisation current comparable with that from the other plants. The test lenses are lacquered indentically to those going through the other machines. Is there anything other than the ion source that could be to blame? There are no alarms. The ion source is a Lion mark 1, and is generally regarded as small for the size of the coating chamber (Boxer).

The Bayer test, IMHO, DOES NOT REPLICATE REAL-WORLD EYEGLASS HANDLING BY REAL-WORLD AMERICANS.

The tumble-drum test from Colts lab, I think, is soo much better in this regard.

Barry

[chip anderson]

Anyone ever performed a siding around on the dash test?
Anyone ever performed a staying in the car under the visor test in the South in Summer?
Anyone ever performed a burning limbs, or BarBQue test?
Anyone ever performed a working under cars all day every day test?
Anyone ever performes a being on the face of a salt water charter boat crewman test?
Working in a pick-up bed coating company?
Working on a construction job?
Working in a sand blasting job?

Not all patients are society ladies whose worst problem is hair spray and make-up removal.

Chip

[Falstaff]

Thanks for the replies. It is true that we might be attaching too much significance to one or two tests, that maybe do not relate to the real world too well. Perhaps the coatings from this machine are as good as they can be, and there is nothing that can be done to make them better.
There is probably nothing we can do in the lab that can adequately pre-empt the conditions likely to be experienced out there in the big bad world. However, I feel that we have to try to do what we can, and continuous improvement is a worthwhile goal.
Our testing regime is based around Colt's methods, which have always served us well; but we must not lose sight of what the final customer experiences. This is the most important test of all.

[Mauro.Airoldi]

Some suggestions:
1 verify vacuum integrity of boxer plant by a leak up rate (it must be less of 5*10-5 mb*l/s)
2 I suppose you use a process "similar" to the Leybold Brilliant, so, verify the correct evaporation rate (sweeper, power...) in all materials and is specific on ZrO2.
3 Are you sure that the neutralization current is negative with argon? (it must be negative with a value of 5-10% of the anode current)
4 If you use Al2O3,Ti3O5 Or Ti2O3 , try to rebuilt the lineer, if it is too old the Bayer give bad result.
in any case consider the Bayer resul reproducibility with a variation of +/-1 and hight result you have and hight is the variation

ps 5 How is the purge valve in rotary pump? if it is too open the pumping speed change, the final vacuum look ok but you haven't the same performance.
let me know

[Chris Ryser]

I have some knowdledge on formulating hardcoats chemically and have no idea about the vacuum coated.
All the testing is for the birds because hardcoat can be made nearly as hard as a diamond.
However if the substrate is soft a really hard coat might just not flex along and is not perfect. CR39 is much harder than poly so the need different grades of hardness. Thin lenses can flex and the coating can craze.

I am wearing a "Radio Jubilee" watch for the last 15 years and it is on my wrist 24 hours per day and still does not show one minor scratch. This prooves that it can be done and could be done 15 years ago, which is a long time in this age of technology.

Totally unscratchable and the new one ones are made with ceramic cases and wristbands. Has anybody seen frames made with the new ceramics. Indestructible. unscratchable, expensive but ever lasting in brand new shape.

[Jeff Trail]

I would lean towards 1 or 5 from Mauro Airoldi post.. one question are you running the tests in-house or a central processor? In house it maybe be your testing that is the culprit ..

One thing to remember all of us tend to lose sight of the forest blocked by a tree.. is it way off or in an acceptable parameter, I know when I was in an AR center we sometimes let the numbers drive us crazy and when you stepped back and looked at the "big picture" it was ok.. we all would have liked every run have 100% transmission but when we started to tinker with the gun /sweeps/ pressure /oxygen and on and on those 10ths started looking and feeling like the grand canyon ..

Guess I am getting more mellow with age :-)

Jeff Trail

[Mauro.Airoldi]

dear Jeff
to explane my suggestion i've to explane than I was AR process engineer in Satisloh for 10 years...
so let me say that there are some "standard procedure" to follows, I've only suggest the simple one.

[Falstaff]

Hello,

Thank you for your help. I have conducted the tests myself, and there is definitely a problem, but it is not there all of the time. All of the tests are done with identical lenses, all hardcoated with the same lacquer. This means that we can get a direct comparisonon on the performance of all the plants.

The Colt's 'eraser' test sometimes removes, or partially removes, the reflex colour on lenses from this plant, which never happens on lenses from our other plants. Spectroscopic analysis of a lens damaged by the eraser test suggests that the final SiO2 layer has been removed, or partially removed. Bayer tests are always less good on lenses from this plant. There has never been an issue with adhesion or crazing however.

So far I have not found any obvious issues with the steps 1 to 4 as suggested above, though I am not sure how to check the valve in step 5.

[Mauro.Airoldi]

"How is the purge valve in rotary pump? if it is too open the pumping speed change, the final vacuum look ok but you haven't the same performance".
if you have a Boxer, rotary pump has a black lever in front, the lever must be un open (if i remember it is right side), in some plant leybolt put a extra valve BEFORE the pump (on the piping from the chamber) you can try to close it.
In any case, what is your leack up rate?
the right procedure is the follow:
clean plant
run hight vacuum for at list 8 hours (NO meisneer select)
Close vacuum valve and, if your plant is a turbo pump, wait for 60 minutes (if it is a Cryo or diffusion pump you must non wait)
keep pressure in chamber P1
wait 10 minutes T1
keep pressure in chamber P2
Chamber volume see on manuals V
leak is: V*(P2-P1)/T1 in millibat*liters/second
lower is value, it is better
Max admitted 5*10-5 mb*l/s

We don't use eraser test like standard because we consider it too "soft" test, if you failed it, you have BIG troubles.
I don't know what process are you using, but if it is a leybold process... before last SiO2 you have a Al2O3 layer... (it is +/- the standard sequence for leybold). So i suggest you to verify Al2O3 layer, replace material in the lineer and try again.
for a better adhesion betwen Al2O3 and SiO2, you can use a extra help, it can be implement a extra layer of IAD (60'') V anode 120-160Vdc, I anode 0.6-0.9 A, I neutralization -0.6-0.1 A.
But let me give my feeling (30 years in vacuum depositions), you have a trouble in vacuum system.
let me know

[Falstaff]

Hello,

I think you may be right about the vacuum. We did a leak tast last night and found the rate to be 1.3 X 10-4mbar l/s. I am currently repeating the test just in case the door seal was to blame.

Many thanks. Your advice is very much appreciated.

Falstaff