French (Fr)English (United Kingdom)

DPC NEWS: a website dedicated to Penetrant Testing and Magnetic Testing



visits on site since April 2008

Log in


Receive HTML?

Low cost but reliable calibrations

Written by Administrator
Thursday, 01 July 2010 10:41

July 2010

Penetrant Testing is more and more dependent on measuring devices: thermometers, radiometers, luxmeters, time-switchers, manometers. What else? As a renowned actor could say in an advertisement for a coffee supplier.

More meters of any kind equals more frequent calibrations, more calibrations, more paper, burden: users need to anticipate when the meters shall be unavailable on the line because of calibration, they have to anticipate orders with the purchase department, they have to make sure the meters will be put back to service as soon as they arrive in the warehouse, etc.

Sometimes it is almost mandatory to have 2 meters, with different “due dates”, so that when one meter is sent to the calibration subcontractor, the second one is put in service for the time needed by the subcontractor - sometimes, several weeks! That means you need a second calibration row for the second meter! More chances to have something going wrong, such as a forgotten due date, a calibration label not affixed on the meter, etc: the more items to be calibrated, the more likely some wrongdoing … that will be detected by an auditor! We all know that auditors have the “feeling”, the “expertise”, to ask for the very minor thing that went wrong, and that is not yet fixed!

Facing such a situation some companies prefer to subcontract completely not only the calibration of all their meters but also the management of the calibration schedule according to the different requirements of standards, specifications, etc. 

Another point to think about is: what if a meter when calibrated is rated as “to be repaired”, making it unavailable for a time longer than anticipated? Will the user have to buy or rent another calibrated meter?

More meters means also skyrocketing calibration expenses.

This paper is to show you that there are means to make reliable and quick calibration which are at the same time low-cost. This is not applicable to all the meters needed during an inspection.

However, at least you may get some ideas to:
• Lower the cost.
• Make fast turnaround calibration.
• Make in-house calibrations.
• Have your penetrant line running for longer times without any stop due to calibration problems.
• Etc.

We are to give you some methods for:
• Thermometers.
• Manometers.
• Time-switches.
• Refractometers.

You are also reminded that a MEASURING device/equipment shall be calibrated, while an INDICATOR shall not!

If you have a look at your penetrant line, you may call some meters “indicators”; hence, no calibration, no cost, no Quality Assurance burden.

One typical example is the manometer which measures the pressure of the air which pushes a product from a pot to an electrostatic spray gun. IN NO WAY, pressure here has any influence on the QUALITY of the process. If pressure is too high, this leads to an over-consumption of products: the materials’ supplier is happy! Too low a pressure has almost no effect except the operator will need more time to apply the product. This the operator will soon notice. BUT the overall quality of the process is not in jeopardy.

Let us go to the practical work!

1- How to calibrate a thermometer at 0°C (32°F) and 100°C (212°F)

Penetrant Testing must comply with many requirements. Among them, several steps require that temperature be checked: temperature of parts as well as temperature of Penetrant Testing materials.

What are these steps?

• A general requirement is that a “standard” Penetrant Testing shall be performed within the 10-50 °C (50-122 °F) range. This includes parts and materials. A thermometer measuring the ambient temperature is the best way: we may assume that, unless the line goes through several rooms separated by walls/doors, the average temperature will be the same in almost every spot of the working area, within 1°C (2°F) from the figure displayed on the thermometer measuring the ambient temperature. See following paragraph.

• Specific requirements are set for:

- When dipping parts in a penetrant tank parts shall be at a temperature of 40°C (104°F) maximum. This is to prevent dipping parts immediately after, say, an alkaline or a vapour phase (this latter being less and less used) degreasing. If they are yet too hot: this could modify the penetrant formula by easing evaporation of the most volatile ingredients. This may be a reason why auditors would check the temperature of the penetrant in the tank all the shift long.

- Rinsing/washing steps, when using water. For instance, some specifications require a water temperature in the 15–35°C (59-95°F) range.

- Drying step: if using a drying oven many specifications allow a drying temperature set at 70°C (158°F) maximum; some even state that the surface temperature of the parts shall not be above 60°C (140°F), or even less.

- Developer application: many specifications require the surface temperature of the parts be at 40°C (104°F) maximum when applying developer, whatever its form (dry powder or non-aqueous wet developer). Water-based developers are applied by immersion of parts in a tank whose temperature is generally set in the range 75-85°C (167-185°F).
A very short story: a manufacturer marketed, up to the end of the‘90s, an aqueous-based liquid developer, which could be also applied by brush or sprayed on the wet surface! After application, the developer layer must be dried in a short time with a flow of warm air. However, this process did not give the expected results.

Reading the above paragraph we understand that Penetrant Testing lines are always run well within the 0-100°C (32-212°F) range, and that we need a digital thermometer using K-type thermocouples for the surface, air and liquid temperature measurement. However, a simple thermometer for liquid may be perfectly used to check a penetrant bath or the room temperature.

We understand also that such thermometers are nowadays used in several steps of the PT process while before the ‘90s this was almost unheard of. These meters shall be calibrated, generally at least every six months as for radiometers, luxmeters and tangential magnetic field strength meters. Calibration may be then performed every 12 months after 4 periods of 6 months if the assurance of the reliability of the calibration process and meter stability is proven. This is a common rule in the American aerospace industry, and it is applied also by its European subcontractors. Contact your NADCAP auditor for more information. Auditors are very harsh about calibration of any measuring device. An easy target, one of the first ones of any audit!

Every one of our readers has a true knowledge of the cost and the time needed for any calibration.

As thermometers would be used, say, between 10 and 85°C (50-185°F), we may think it would be nice to calibrate them at 0°C and 100°C, then covering more than the entire needed scale.

By the way, these figures: 0°C and 100°C … do they mean something for you??

It seems that they are respectively:
• 1- The freezing point of water.
• 2- The boiling point of water.

In fact, this is not exactly true as per physics.

• Freezing point:

Note that the triple point of water, i.e. the single combination of pressure and temperature at which liquid water, solid ice and water vapour can coexist in a stable equilibrium, occurs at exactly 273.16 K (0.01°C) and a partial vapour pressure of 611 pascals.
Therefore, there is just 0.01°C between the melting point and the triple point.

But obviously such a difference needs not to be considered for PT!

• Boiling point:

The 100°C figure is valid ONLY for a pressure of 101,325 Pa (1,013.25 hPa), the standard atmosphere pressure on Earth. But, even if the atmosphere pressure drops down to 985 hPa - a very low pressure - the boiling point will be 0.5°C lower, and if it goes up to 1,040 hPa - a very strong anti cyclonic condition - the boiling point will be 1°C higher.(*)

Once again, we may assume that such a tiny difference would not have ANY DETRIMENTAL EFFECT on PT!

When we are at this point of the paper, don’t you begin to guess how YOU could achieve a good, cheap, quick and reliable calibration of the thermometers of a PT line?

Quality-assurance auditors and specifications/procedures

One of the first tasks of an auditor is to check that the entire PT line, including inspection, is run after the applicable documents. Among the very first documents the auditor will check the calibration procedures, records, the labeling of every measuring device, all in all the traceability of the calibration are very close to the top of the list.

If something, anything, is performed without being substantiated by a document, you will be unease. BUT, if you show the auditor the applicable document, if you show you have done exactly as written in the document, if you are able to show all the records, then you will feel at ease!

So, what about writing your own procedure, taking advantage of the immutable - or so - parameters that every calibration laboratory will use in the calibration chain of thermometers?

If you write a calibration procedure, if you follow this procedure to calibrate your meters, if you allow a reasonable tolerance to the results you get, which auditor could prevent you to use this procedure? Would he be sure that the task would be better done by an outside laboratory?

You may anticipate some advantages:
• Much cheaper calibration.
• Much shorter delivery times.
• Less paperwork, even for the purchase department.
• The savings could be used for more productive use in your department.
• And probably others!

So, what i have to do?

You write a procedure, after the rules, which describes what will be done.

Here are some ideas - very easy to perform, very efficient, very reliable and so cheap!

1- Calibration of the 0°C point

You need to have some ice cubes in a refrigerator. A beaker, tap water (the very small quantity of chemicals: calcium and magnesium salts, sodium hypochlorite, etc. which are dissolved will have no effect on the freezing temperature).
Add tap water to the ice cubes in the beaker. Allow for stabilization. Immerse your probe into the upper layer of water. You are SURE that, by definition, the temperature is 0°C (32°F)!

Your brain could show a reaction: “Wait a minute: in the beginning of the paper, you wrote “freezing point”, i.e. the temperature when water becomes ice. Here you ask us to make the test when ice becomes water. Is there something wrong?”

We must remind you, just in case, that the freezing point of a liquid is EXACTLY identical to the melting point of this same material when solid - as long as the material is pure. And we are right in the good condition!

2- Calibration of the 100°C point

With the same way of thinking, you will have a beaker atop a Bunsen gas burner, or an electric heater item. You pour tap water to the beaker, make it boil, and you immerse the sensor into the top layer of water, or just atop. You then have the 100°C mark. Right, it should be adjusted, depending on the real atmospheric pressure in the laboratory. But the difference is so small that, once again, we can have no consideration for it, due to the needs of temperature measuring on a PT line.

Nevertheless, we suggest that, when audited, you have available a table, or a curve, showing the very, very small influence of the atmospheric pressure in the laboratory on the real value of the boiling water. Just in case the auditor would be suspicious.

3- Results

PT is not that dependent on a difference of a few °C. Hence, you could write that your calibration is right +/-1°C (2°F).

4- Recording of data

Your procedure must detail how data are recorded, which means you use to trace the data and the meter (a label on the meter, for instance, with the certificate number, the date of calibration and the due date for the next calibration). Some modern digital thermometers offer USB data output and Bluetooth® wireless technology for cable free data transfer to a PC.

No auditor can prevent you doing such calibrations.

When we see how some radiometers and some luxmeters are calibrated, without any question by the auditor about the process, we would be happy to help you if you meet some reluctance from some badly-informed auditors.

(*) The boiling temperature of water for atmospheric pressures is got through the Antoine equation:

Where the temperature Tb is the boiling point in °C and P is the pressure in torrs.

1 hPa = 0.750061683 torr.

Details of calculus:

2- Manometers

This paragraph cannot help you in the calibration process. It is only to give you a VERY PRACTICAL way to lower calibration costs, to reduce the line time-out during calibration.

Many users think that:

• Calibration shall be performed by an outside company, specialized in calibration. Why not?

• As an alternative, calibration could be performed in-house, in the calibration department. This department has a “reference manometer”, calibrated by an outside specialist. Nevertheless, the manometer shall be sent to the in-house department. This shall be preferably scheduled when the PT line is not in-service: week-ends, bank holidays, maintenance times, etc.

We are close to the cheapest way of performing the calibration. Why would users have to send the meter to the calibration department, meaning disconnecting the meter, emptying it of water, if any, then reconnecting it, doing what is needed to prevent any leak?

What if the calibration department guy comes with the reference manometer along the line and performs the calibration there? Without any dismantling?? How is that possible???

Just have a T-plug on your manometer!!! One of the branches goes to the manometer, the “empty” one is ready to be connected to the … reference manometer! Doing so, you lower a lot of burdens, risks about leaks, no risk to “overtorque”, etc; the line is stopped for a very short time, traceability is assured as long as the calibration department can show any auditor the requested documents.

A nice idea, isn’t it?

3- Time-switches

Time-switches are used on almost every step of a PT process. Some time-frames are short, 30 seconds, others more in the dozens of minutes. That means that the measuring devices shall be calibrated. Several devices!! Costs may be impressive, especially if, once again, if the “meter” is to be sent outside, this will need such a long time (we are talking of days, even weeks) that a replacement is needed on the line. Calibration costs are easily doubled! Paperwork also, even if nowadays “paper” means “computer” … but many know that it is easy to spend more time with a restive computer to perform a simple task than by using “real” paper!

What is described now is a VERY EFFECTIVE, VERY CHEAP AND VERY RELIABLE means to calibrate time-switches in-situ. The effectiveness/cost ratio is incredible!

Do you know the time-measuring equipment which has an accuracy of one second per million years? An atomic clock.

An atomic clock? Are you foolish? It is very expensive, it is large. How, as an average person, could I have access to such a tremendously precise equipment?

In many countries, there is a phone service - computer-literate people would say “an application”!! - called the “Speaking clock” or similar … which is connected to the “time reference”, a world network which allows for this unique accuracy.

So what we suggest is very easy and very cheap: you dial the specific number, you hear the voice, then you make your time-switch start when you hear, say, one minute and zero second. Then depending of the time span you need (for instance, up to three minutes, for the emulsifying step) you wait for, say, 4 minutes, so as to have a calibration point beyond the longest time you will ever use in this step. If the time-switch is used on the penetrant tank, where you have a dwell time, say, of 15, 20 or 30 minutes, depending on parts and process, you disconnect the phone line, say, for 14 minutes, and you wait for the “15 minutes” mark (i.e. the 16 minutes zero second “beep”). And so forth for the other times. You record the time you have seen when the voice says 16 and zero second and says 16 minutes and 2 seconds. Same process for the 20 and 30-minute marks.
The purpose is to have a measured point beyond the 15 minutes, the 20 minutes and 30 minutes in order to calibrate beyond the range normally used.

As far as a procedure describes this, which auditor could state your reference is not right? We may guess that for 30 seconds time-span you will have no difference between the time-switch and the “Speaking Clock”. And if there is a second, or two, or three, or even five, on the 30-minute mark, who cares? It is RECORDED: your meter IS CALIBRATED!

We can tell you that many companies in Europe do this way…and it works perfectly! No auditor, as far as we know, has EVER refused this! We assume we would have been told if it were the case!

4- Refractometers

Imagine you have a meter, whatever it is. You send it for calibration to a National calibration laboratory.

It comes back calibrated, and some days later, the purchase department receives the invoice. It is redirected to you, as you must, being the people who ordered the service, okayed it. When reading the amount, you think there is a mistake: the cost for calibration is … 3 times the price of a brand new meter!!

Unfortunately for you, no mistake: that’s the cost!!

That is the situation about refractometers. So you wonder what to do the next time: buy a new refractometer every year? Find an idea?

Hurrah! The idea is here, in this paper! Once again, far cheaper than the calibration by a National laboratory. And the calibration will MEET YOUR NEEDS, NOT MORE.

A refractometer is needed when you use an hydrophilic emulsifier. Also if you use a water-based penetrant.

We are to give you an idea for writing a procedure. You need some glassware, and a person used to chemistry laboratory job. An engineer level is NOT mandatory!

Guess you use an hydrophilic emulsifier diluted at 5% in water. You need to calibrate your refractometer between, say, zero and 7%.

The zero mark is easy to get: the water which is used to dilute the hydrophilic emulsifier! It may be tap water, but also may be an “industrial-quality” water. You must use the water which is used for the hydrophilic emulsifier dilution on the line. This will give you the zero reading, and generally you can adjust it through a very small screw at the bottom of the unit.

We suggest you prepare one liter (ca a quarter) each of several concentrations of hydrophilic emulsifier in water, as suggested underneath:
• 1%
• 2%
• 4%
• 4.5%
• 5%
• 5.5%
• 6%
• 7%

This method is based on the measurement of refractive index of liquids. Since it varies with temperature, measurements should be carried out at the SAME temperature which MUST be written on the calibration graph, except if the refractometer is automatically temperature-adjusted: have a look to the supplier’s instructions or technical data sheet.
If not automatically adjusted, this calibration graph may be used to check in-service hydrophilic emulsifier solutions provided that the temperature of the sample is within +/- 3°C (+/-5.5°F) from the temperature when calibrated. This range MUST be also written on the calibration graph to avoid any dispute.

For example, if the temperature written on the graph is 20°C, it means that this graph may be used for samples in the 17-23°C (63-73°F) range.

Why one liter, when only one drop of the solution will be used for calibration? This is to lower the relative errors when diluting. If you dilute 1 mL in 100 mL, you have to use a 1 mL pipette. A tiny 0.1 mL error translates into a 10% relative error. When using a 10 mL pipette for one liter, should there is a 0.3 mL mistake, this comes to a 3% relative error.

To prevent a slow drift of the solution (which must be kept tightly closed in the dark and at ambient temperature), solutions are made once a year.

Once again, calibration means - and cost - must be connected to the needs. What we have described meets the needs of PT users.

We will end this document with an example which is to substantiate the last two sentences.

We know of the subcontractor of a large prime who was requested to use a manometer on the water rinsing spray lance of his PT line. Nothing unusual.

What made us scratching our heads is that the auditor demanded that the manometer had a resolution of 1/100 bar (i.e.: 10 hPa; 0.15 psi). A very costly device, costly to calibrate … for which purpose???

So before accepting any request by a prime, an auditor, a purchase department, try to think a bit. We have seen so many mistakes, for instance, in documents when converting US units in SI units!! Sometimes it becomes ABSOLUTELY impossible to meet the requirements! When you are requested that your measuring micrometers must have a resolution of 0.00001 mm … i.e. 10 nm … do you have to comply? The answer is a definitely vibrant: NO!

Last Updated ( Wednesday, 25 May 2011 11:05 )