Field Hardness Testing - Experiences? 2

[Guest102023]

I am looking for people's experiences with field hardness testing on boiler materials, including tubes and including HAZs.

In particular I am thinking about the Microdur MIC-10 tester, whose accuracy is adversely affected by workpiece vibration effects. What schemes are there for stiffening up thin-wall and/or small-diameter piping/tubing for Microdur testing? What are you using for calibration blocks?

Anyone have experience with the optical through-indenter-viewing (TIV) method? It sounds very promising.

 

[Metaljon]

We use the MIC10 hardness tester. It is important with this type of hardness tester to use hardness blocks for calibration of the instrument. We try if possible to use a similar material for the blocks and we have prepared are own standards (none commercially available for the range of materials we require). We have carbon steel, P22, P91 and 316 hardness standards which we test and certify ourselves. We also age the calibration blocks so that we validate the readings over a reasonable hardness range from virgin to degraded. During site jobs, we test the calibration of the instrument at least daily. The GE software for this instrument is poor and we have developed are own macros for this work.

As you point our the MIC10 instrument is prone to external vibration and has limitation on thickness. If the boiler tube lengths are long sometimes we clamp the tubes to increase rigidity.

We normally measure weld and parent metal hardness. We don't routinely record HAZ as it is a very unreliable test. The HAZ is composed of a range of microstructures and this will cause variation in hardness. Sometime for petrochemical work we measure HAZ but I always qualify the reliability as being low.

The optical method TIV is not suitable for site work in my opinion but ok for laboratory work.

For pipes, vessels and headers we use the latest Equotip 3 instrument. This gives excellent results provided that the instrument is set up and calibrated properly. It allows the calibration curve for a material to be programmed into the instrument which gives very consistent results +/-2%. Excellent software output for Equotip 3 and makes the reporting a breeze.

As with all hardness testers the MIC10 requires a very good surface finish. I have seen some contractors using rough finish and the results are poor. Preparation, calibration and training are key elements to get good results with this instrument.

 

[metengr]

Regarding boiler tubes or heat exchanger tubes, I have very mixed feelings regarding the use of any of the current portable hardness testers for field application. My opinion is that in-situ hardness testing of boiler tubing is unreliable, at best.

Pipe and plate end products are a different matter regarding portable hardness testing, as mentioned above.

 

[Guest102023]

Metaljon,

Thanks, but why would the TIV not be suitable, provided workpiece access is good?

 

[Metaljon]

We have tested the TIV and in my opinion it is not practical for onsite measurement of hardness. The probe is very sensitive to probe alignment and surface finish. If these are not well controlled, the TIV gives quite large scatter in hardness. UCI method also gives scatter but is less sensitive that TIV. TIV is similar to a micro-hardness tester. As the indents are small the technique is highly surface sensitive and any abnormality in surface finish will affect the results. Also the stability of the sample needs to be considered as this will affect the load being applied. Despite this I think TIV has some use in a laboratory setting for coatings and thin materials where test parameters can be tightly controlled. If you review the GE documentation you will see the bounds of applicability for TIV. Under tubes, casting and forgings it lists the technique as being "partly" applicable. I don't know what "partly" means but I think in marketing terms it means "probably not suitable".

 

[ak1965]

brimstoner


to me... for field usage.... poldi hardness tester givsthe best results....! looks bit old fashioned.. but very very accurate.... you know ... OLD is GOLD..!

All electronic hardness tester whether MIC 10, Equotip have some or other limitations..!

All of the electronic testers give good results when the strucutre under test is heavy ... otherwise.. what you said is right...vibrations create hell of the confusion...! we all konw that calibration pieces are made heavy .. we get good results..!

SO try POLDI ... and rely your own eyes...!

 

[Guest102023]

Metaljon,

I anticipate working on the Cr-Mo alloys in addition to carbon steel (with the MIC-10), so my next step is obtaining calibration standards in 2¼Cr-Mo and 9Cr-Mo. How do you do it and ensure consistency? Are they available ready-made?

I am still a little puzzled by MIC-10's insistence on 'similar material' for the blocks ... I should think that similar class (steels incl. low alloy) and the same elastic modulus are sufficient. I think if anything the YS:UTS ratio should be more significant.

 

[Mr168]

EPRI just had a recent conference over this issue a few weeks ago in their Charlotte branch.

I'm hoping they send out copies of the first day presentations, which were available to the public. They had compared the readings from the various methods and common equipment, including standard deviations and inaccuracies between personnel working at the same labs.

Needless to say, it was very eye-opening, and not particularly confidence-inspiring when forced to rely on hardness measurements as a key source of damage analysis. If you do some fishing, someone may already have the notes.

 

[Guest102023]

I am eagerly anticipating that report!
My personal estimate of the percentage accuracy of field hardness testing is continually slipping downward; it is now officially below 10%.
And that figure excludes Brinell 'testing' done by heat treat contractors ...

 

[Metaljon]

It is true that the general accuracy of the portable hardness is 10%. However if care is taken with the calibration of the instruments using standards, training of technicians and good surface preparation, the accuracy achieved is much better. We regularly achieve accuracy of better than 2% and that is for difficult materials like P91 and stainless. We also check using a pin brinell hardness tester onsite and the results show a very close match with the portable hardness testers. The standard programs that come with Equotip and MIC10 are not good with P91 or stainless. The latest Equotip3 instrument allows input of best curve fit and this gives excellent correlation for aged P91 and stainless when compared to pin brinell. Also it is important to select the appropriate hardness instrument for the job. We use Equotip3 for reasonably thick wall, MIC10 for thinner wall and tubes and TIV for the super thin wall/coating applications. Equotip3 is best for castings and forgings and we don't use MIC10 for this application. I find the biggest inaccuracies is from technicians who don't understand the limitations of the equipment and believe the handbook supplied as being the whole truth and nothing but the truth. The handbooks supplied are only a guide more like marketing literature. I find best to get metallurgists to do hardness testing as they have a better understanding of the material they are testing. However these days every man and is dog has a hardness tester which is why the accuracy of the equipment has fallen into disrepute.

 

[Guest102023]

Metaljon,

I concur with your first statement; in my experience the MIC-10 has rarely been better than ±10%, but then the major chain I worked for did do training beyond turning the instrument on and calibrating. However I think in the field it can be useful for finding glaring errors caused by incorrect PWHT of P91, for example.

I could not agree more with your last couple of sentences.
When you combine the general tendency of semi-technical types and laymen to believe anything that gets displayed on an instrument with the dumbing-down policies of inspection companies, the test result is usually expensive garbage. I see the same thing with PMI, where operators don't understand that most of the displayed digits are random numbers, not significant figures, and it then gets called 'chemical analysis'.

[We need a separate thread dedicated to the practices of chain inspection companies (I will declare my interest now, since I compete with them in a few areas). But we are also considering equipment reliability and worker safety here, so I am not entirely self-interested.]

 

[vesselguy]

Metaljon,

Pardon my ignorance but I would like to know what do you mean when you say "pin brinell"? What is that? You don't mean doing microhardness testing such as Vickers, do you?

 

[Guest102023]

http://www.westportcorp.com/hardness/brinnell/PinBrinellTesters.asp

 

[Duwe6]

"Pin Brinnell" is an old-style tester that you whacked with a hammer hard enough that the pin broke, providing a 'calibrated' impact force. Indentation is then read with 'scope like classic Brinnell.

Other than for side-by-side comparisons of materials, "Rebound" testerl like the Equotip are basically worthless in the field. Looks good on the huge, polished, dead-flat cal block. Gives irreproducible results in the field. You pick a hardness, and I can just keep testing until that number is found. But it won't be repeatable.

The only tester with reproducable and accurate results I have used, including the Mic 10, is the "Telebrinneller" style. As long as the holder is hit hard enough to give an approx. 3mm indentation in the workpiece, it is pretty accurate. And if you are looking for "less than 200 BHN", if you use a 200 BHN bar for the reference bar, all you have to do is 'eyeball' the indentations. If the indent in the work is noticebally larger than the one in the bar, you write "significantly less than 200 BHN". No squinting in the microscope required.

 

[Metaljon]

The pin brinell is very useful for verifying the accuracy of the portable hardness testers. Just needs a consistent hammer blow and then measure the indent. It has a shear pin that must break otherwise reading is invalid. Very easy to use. Only downside is that it leave a big brinell indent in the surface and it is a bit slow.

We manufacture our own hardness blocks as there is nothing commercially available. We use reasonably thick sections approx 40mm plate. We have our own heat treatment furnace in the lab. I normalise the samples following by quench and temper cycles. I temper back into about 6 conditions covering the whole range from quenched to very degraded condition. For example for P91 we have hardness blocks ranging from 400HB to 160 HB. We check the hardness using a brinell tester 3000kg and then compare the results to the portable hardness readings. We then do a curve fit and find best polynominal. For Equotip3 it gives an option to input curve fit parameters. For MIC10 we store the data and then re-manipulate data using best curve fit. We check hardness onsite using pin brinell. We always get very consistent results. Also we check our hardness tester daily to check for no drift.

The problem with the MIC10 instrument is that it assume a linear interpolation across the whole hardness range which is not correct for P22 and 9Cr. The standard program supplied with the Equotip3 for steels or tools steels does not work for 12Cr, P91 or aged stainless. This is why you have to calibrate the instruments and setup yourself. If you use the standard program setup for MIC10 and Equotip3 then expect an accuracy of about 10%.

http://nz.linkedin.com/pub/jonathan-smith/39/25/589