Help solving hardness measurement difference

[learnfast]

Hi, I am having hardness (both RC and R15N) measurement difference with my heat treating supplier. My measurement is usually 1-2 points higher than my supplier's. We did a hardness correlation study with 15 parts engraved with numbers so that everyone measures the same part. My average is RC 54.5 and R15N 87.6. The heat treating supplier average is RC 52.8 and R15N 85.9. The heat treating specification is RC<54.

We are all trying to understand why this happens since we already tried to standardize everything (anvil size, measurement location).

Any input is highly appreciated.

Thank you

 

[swall]

Your results are consistent with what I have experienced with "round robin" hardness testing.

 

[TemperedMartensite]

I would say skip measuring parts and start using callibration standards. Use the same 15N or HRC block at your shop and at the heat treaters.

If you really want to get an idea you could use a microhardness standard and check it on both pieces of equipment. Then you verify the measurement by doing a knoop/vickers test next to the 15N or HRC measurment.

Why are they different? Now that is a tougher question to answer, but the mechanisms for applying the loads may not be exactly the same.

You mention you have standarized the anvil and location...
but are you using the same types of eqipment (mechaincal vs. load cell)?

 

[redpicker]

I'd tend to agree with swall. Your HRC numbers are just 1.7 HRC points different. Considering that standardized test blocks are only accurate to within +/- 1 HRC point, you can't say that this difference is significant.

That is, if you had a standardized test block that was certified as 43.7 HRC +/- 1 and got these results on both machines, they both would be considered acceptable.

Standardized test blocks are generally made to tighter controls on the uniformity of the hardness as well as surface finish than routine parts, so the differenct you are seeing doesn't indicate there is anything wrong.

rp

 

[TVP]

I agree with the previous comments, and suggest you try the same calibration block on both testers (if possible) in order to see how much variation is in the machines/method.

 

[EdStainless]

Did you analyze this as a true Gage R&R?
You need to do that.
Be rigorous on the method and statistics, then you will know if it is significant.

I have never seen a hardness study that reported they could do any better than RC 50 +/-1.5. It is the nature of the beast.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[blacksmith37]

And, the hardness test equipment should be calibrated/standardized (usually each year) by an independent party ; there should be a sticker or tag on the machine with date, etc, of the last calibration.

 

[learnfast]

Thanks a lot for everyone's input. I used MiniTab to compare the data and found there was significant difference between my measurement and heat treating supplier's measurement.

I am going to send standard hardness blocks (RC and R15N) to my heat treating supplier to see if we have the same readings.

Both sides all calibrate hardness tester with a standard block everyday. And both sides hardness tester are calibrated by third party every 6 months.

I will also check if the same types of equipment (mechanical vs. load cell) is used.

Have a nice weekend

 

[CoryPad]

Try to "hide" the test block actual value from the heat treating supplier. Technicians sometimes influence the results to make the results match the test block values.

What is the hardness of the calibration block used every day? It is best if it is close to the values you are trying measure on your parts.

 

[bruv]

Rather than sending the test blocks to your supplier, can you pay him a visit? That way you will be able to see what is going on, check his calibration, see if histest blocks are old and worn out, ensure that he is following his own procedures, and hide the certified values of your test blocks. Following on from that, you might wish to invite him (and his calibration blocks) to your facility to carry out the same set of tests.

 

[jsokal]

One thing that many people do not respect and may even be the cause for the differences in readings is the proximity of the hardness indentation to a neighbouring one. A good rule to remember is never take a hardness check closer than 3 times the diameter of an indentation. If you are too close to a neighbouring indentation, you will get an artificially high reading.

John

Denial ain't just a river in Egypt...Mark Twain