Spark Testing for Coating Applications

[energy]

Does anybody have experience in calibrating a High Frequency Generator Tester for Leak Detection? We have an Electro Technic Products Model BD-60 (2,000-40,000 @ 1/2 Megahertz Frequency) that is used to detect pin holes in our Pressure Vessels interior lining. My question has to do with the Calibrator we purchased to determine the proper spark length (gap). We will be, undoubtably, be questioned on the calibration of this device by our future Registrar Auditor. The instructions that came with the Calibrator has caused me some concern as to the proper method for determining the actual voltage applied to the coated surface. Any help will be appreciated. I figure I would try here before getting into it with the Manufacturer. You know, practical hands-on experience?

 

[Jerry Eldred]

I don't think I'll be of much solace. I don't know anything about the particular manufacturer of the unit. But my experience has been that manufacturer's calibration instructions range anywhere from excellent to nothing.

In the cases where there is apparently nothing, there is no substitute in my experience for doing the digging. you'll have to pull out the claimed specifications, then spend some time on the phone with them digging out their method for making the measurements traceable.

There is the possibility in some cases that manufacturer's won't be forthcoming. The best you can do in some cases is get their method good or bad, and use that method. If the traceability seems inadequate, there is a grey line somewhere between doing it the same way the manufacturer does... period; and determining the manufacturer's method simply isn't adequate.

I took a look at their website. I didn't see any plu/minus tolerances listed. Some of the issue depends on how tight they spec the units. If it is simply a range of voltages, it isn't uncommon to see a spec of < for the bottom end to > for the top end. If that is the case, the required method won't be any more stringent than the spec (which in my example is very loose).

I think the test method would probably need to be a high voltage probe with adequate bandwidth and voltage. I'd be curious to see the calibration instructions. If you have something you could email to me, I may be able to give you some more detailed comment offline.

 

[energy]

Thanks, Jerry

I kind of thought so. You're interest in this is appreciated. The fact that you went to their website looking for assistance, impressed me. I will scan the instructions and send them to you along with what I see as a problem. Please stay tuned.

 

[energy]

What, me Worry?

Jerry,

The concern I had regarding the amount of voltage applied to the lining of our vessels has diminished considerably since I obtained the Operating Sheet for the Calibrator. The instructions are pretty much what you alluded to in your post. Statements such as “may vary between 200-400V per Mil of thickness” and “to adjust sensitivity, reduce sensitivity somewhat” and “there is approx. ¼” spark difference between starting voltage and breaking voltage”, indicates that there is no exact voltage measurement possible between coating applications. It would drive someone with a desire to “measure” actual voltage, batty. The calibrator gets you in the ballpark with the generator settings. So, I can have the machine set to deliver a voltage between 2 to 4kV for a 10 Mils of coating. 4kV-8kV for 20 mils and so on. We typically apply a minimum of 60 mils up to 80-100 mils. It’s a spray and bake operation. You can do the math. One good thing came from this, besides peace of mind. Whoever wrote our work instruction saying we apply 6000-8000 Volts to our lining must have only read the first sentence, or were out to lunch. As I was simulating the spark gap we usually use, the calibrator was indicating approx. 25,000 Volts. I was verifying the procedure, before releasing it. We will put in the voltage ranges we may be obtaining in the procedure and simply scribe a reference mark on the power setting and the sensitivity setting to make sure that the knobs are in the same location all the time. (approx.) Once a quarter, we will verify the settings and ranges and document it. As for the Calibrator, I will probably send it out to the Manufacturer every two years for its own verification. Maybe three years. The use for this device is really to make sure that you are not applying excessive voltage/corona that could damage the lining and enough voltage to ensure that there are no pinholes or bare spots. There is a wide spread between the two parameters. Thanks for the response. If you see anything here that you think would cause an Auditor concern, let me know.
And, yes, we have self calibrating Coating Thickness Tester.