Rockwell test and surface finish

G

gammarad

Joined
Jun 15, 2015
Messages
74
Maybe it's a basic question, but is it required to have a certain surface finish to get accurate Rc measurements? I like to measure as soon in the process as possible, just in case I have to re-treat, and that frequently means right out of the coarse 40 grit belt... but today I was wondering if that rough finish may affect the measured value in any significant way.
 
It needs to be flat and smooth. A quick spot touch with a cratex wheel or a 400 grit belt will work.
 
Last edited:
Definitely, if you just think about how shallow the measured penetration is. The allowed surface roughness should be negligible in comparison to the penetrator mark.
Also to consider the other side of the sample. If reamarkable roughness is present in the specimen/anvil interface you start asking too much to the 10 kg preload to squeeze and settle things up for the test to be reliable.
 
Yes the finish should be 400 grit or better ,sides parallel, firmly supported .
 
I take mine to 400 grit. That said before I got my oven and tester I sent my blades out to a couple different places, several times each over the course of 12 years. They had all been tested but none had ever had the mill scale removed or any sanding done before they tested and sent them back.
 
gammarad said:
Maybe it's a basic question, but is it required to have a certain surface finish to get accurate Rc measurements? I like to measure as soon in the process as possible, just in case I have to re-treat, and that frequently means right out of the coarse 40 grit belt... but today I was wondering if that rough finish may affect the measured value in any significant way.
Click to expand...
I wouldn't heat treat at 40 grit,, much less test hardness.
 
Ryan Minchew said:
I take mine to 400 grit. That said before I got my oven and tester I sent my blades out to a couple different places, several times each over the course of 12 years. They had all been tested but none had ever had the mill scale removed or any sanding done before they tested and sent them back.
Click to expand...

Stacy (and rest of ya'll knowledgeable folks) - is there any way to get a valid test without removing the mill scale as described above? Ryan, please understand this is not questioning your work, you just reporting what you got back.

Ken H>
 
Ken H> said:
Stacy (and rest of ya'll knowledgeable folks) - is there any way to get a valid test without removing the mill scale as described above? Ryan, please understand this is not questioning your work, you just reporting what you got back.

Ken H>
Click to expand...

Trust me I always thought it was weird. Even if I had a place sanded off it was never tested. I never grind bevels before heat treat. It was always the same steel. Every edge was tested and I never had any issues, as they were reputable and known heat treaters. But it went against everything I'd read.
 
Ryan - I'll bet what happened is the HT'er knew the steel, knew his process and "knew" what the resulting Rc "should" be and just reported that. OR - could you see any "dimples" left from the HT'ing in the scale?
 
There were always 3 dimples from the tester. I assume like you the treater knew the steel inside and out, but I still always test each blade. I never had any complaints and never any suspect blade edges.
 
When I send a batch of 50 blades to Peter's, they may not test every one, but there will be a polished circle with a divot on the tang of most of them. Every separate group of steels is tested and the report noted on the invoice.

You don't have to remove all the scale to test, just remove it in the spot the test will be done. I use a coarse cratex wheel on a bench grinder.. It will quickly grind a place for the test. A long as the other side is solid and flat, you can leave the scale on there.
 
Thanks for the info, guys. I think it makes sense to assume the roughness of the surface needs to be negligible vs the depth of the divot.

Stacy, I'll give the cratex wheel a try. It sounds like the simplest solution.
 
You must log in or register to reply here.
Back
Top