Three Questions about Micrometers

[CHI_Tool&Die]

It would be difficult to check a lot of the dimensions I have to at work just using an indicator and gage blocks. In other cases it would be complete overkill. Plus using an indicator with blocks isn’t necessarily any more accurate since you still have plenty of opportunities to alter the measurement once moving the indicator and stylus. I have no worries about my mics holding .0001”. That is exactly what they were designed for. If you were that concerned, wouldn’t you use CMM?

 

[Mgdoug3]

If you want to measure engine parts, like a crankshaft, IMO you should be using .0001" graduation micrometers.

I questioned this until I looked at a John Deere manual and it does give a range to 4 decimals. Also remembered the current Massey Ferguson engine I'm overhauling and the total range is 0.007".

 

[X1 Mike]

It would be difficult to check a lot of the dimensions I have to at work just using an indicator and gage blocks. In other cases it would be complete overkill. Plus using an indicator with blocks isn’t necessarily any more accurate since you still have plenty of opportunities to alter the measurement once moving the indicator and stylus. I have no worries about my mics holding .0001”. That is exactly what they were designed for. If you were that concerned, wouldn’t you use CMM?

It may be difficult, but you can bet that if it's an uber critical dimension there is a way.

Yes, CMM makes everything easier but it's not available everywhere and isn't quite so easy to use while machining. They put men on the moon without it so you can still do everything with jeweled instruments.

Being in tool and die you understand die breakage or clearance. General rule of thumb is 10% of metal thickness is the difference between the punch and the die. Then think about something as simple as a razor blade. You make a die for a razor blade and that razor blade is probable between 0.010 - 0.012" which leaves about a grand clearance. The more parts you start lining up together the error keeps adding so you can see how using a mic would be a no go.

My whole point is that they do have a scale that reads to 0.0001" but it cannot be used for anything other than reference. Just because I can put an eye loupe in and scale a dimension within 0.005" I damn sure aint going to stamp it with my name.

 

[lilscorpion]

Whichever route you end up going I would also recommend buying gauge blocks. I occasionally check my micrometers and calipers to make sure they are accurate.

Woha?!? I have 3 sets. I have never ever checked any of them. Ironically all parts are the "same" no matter which pair I use. Been that way since I bought them between 15 and 20 years ago.

Think the resolution of calipers is usually far more accurate than the parts any of us are making even if some of us are in fact machinists...

#allingoodfun

 

[RoninB4]

Measuring instruments may "read" to .0001's but should it be trusted? I have worked in .0001's a good portion of my shop career and working in sub .001's should not be considered a casual affair. Yes many of us have done the +/- .0005 target but to get a reliable dimension closer than that brings temperature, the instrument, and who took the dimension into question. A DTI, air gauging, gauge blocks, or an indicating mic is all I'd trust if I really need to work to +/- .0001. Nice to have a mic that reads to it but holding the mic or the work in your hand can/will change the dimension. Working to +/- .0001 isn't easy even on a surface grinder in good condition. I don't think most people realize just how tiny .0001 really is, let alone measure it reliably. YMMV.

 

[X1 Mike]

Measuring instruments may "read" to .0001's but should it be trusted? I have worked in .0001's a good portion of my shop career and working in sub .001's should not be considered a casual affair. Yes many of us have done the +/- .0005 target but to get a reliable dimension closer than that brings temperature, the instrument, and who took the dimension into question. A DTI, air gauging, gauge blocks, or an indicating mic is all I'd trust if I really need to work to +/- .0001. Nice to have a mic that reads to it but holding the mic or the work in your hand can/will change the dimension. Working to +/- .0001 isn't easy even on a surface grinder in good condition. I don't think most people realize just how tiny .0001 really is, let alone measure it reliably. YMMV.

Along the same lines of what you are saying is that one of the reasons I went into the surface grinding department is that they were "climate controlled" in other words air conditioned.

When you are a hotshot kid and hog off a bunch of stock on the surface grinder and you think you have a few thou left for finish, but it cools undersize you learn real fast.

I would keep a line up of a few parts that were the same and rough them down one by one, then I would semi finish them and finally finish them. Basically, what I'm saying is that it was a 3 step process to grind 0.010" off and finish a part to hold tolerance. The pieces that weren't being worked on would have the surface that was being worked on a plate of steel to take the heat out. Holding 0.0001" is no joke and it only gets more ridiculous when they give you a 50 millionth tolerance (0.00005").

When you are holding those tolerances, you take the bigger 6" x 12" chuck off the surface grinder and put the smaller 6" x 6" chuck on because you need to regrind it to be absolutely flat and the 6x6 is going to heat up way less than a 6 x 12.

So true that people have a hard time wrapping their brains around 0.0001" a piece of regular paper is between 0.003 and 0.004" it's hard to comprehend that the piece of paper is 30 - 40 times thicker than 0.0001.

 

[CHI_Tool&Die]

It may be difficult, but you can bet that if it's an uber critical dimension there is a way.

Yes, CMM makes everything easier but it's not available everywhere and isn't quite so easy to use while machining. They put men on the moon without it so you can still do everything with jeweled instruments.

Being in tool and die you understand die breakage or clearance. General rule of thumb is 10% of metal thickness is the difference between the punch and the die. Then think about something as simple as a razor blade. You make a die for a razor blade and that razor blade is probable between 0.010 - 0.012" which leaves about a grand clearance. The more parts you start lining up together the error keeps adding so you can see how using a mic would be a no go.

My whole point is that they do have a scale that reads to 0.0001" but it cannot be used for anything other than reference. Just because I can put an eye loupe in and scale a dimension within 0.005" I damn sure aint going to stamp it with my name.

I totally understand what you are saying, but I've routinely taken measurements on my micrometers and then trammed using gage blocks and an indicator, ran probing ops, and I've also quadruple-checked with CMM. 99% of the time I'm hitting within +/-.00005". So I don't think I'd say that the micrometer is just for reference -- in my case which is a climate-controlled shop running incredibly precise machines. I agree with you that if you aren't used to running these tolerances you should take the time to check beyond using a mic. When I used to surface grind, my heights were always out because of my mic and that is when I learned to just use the blocks and dial indicator.

PS. It's nice to know there are some other machinists on the GJ!

 

[RoninB4]

A bit off topic so ignore if you don't have the time to read this:

First encounter with temp variation was at a facility that made gauges (Spiral Step Tool/Midwest Gauge Lab) among other functions. SOP to pull a piece off the jig grinder/surface grinder and take it into the climate controlled room to let it "adjust" to the ambiance of the room before deciding if it was good or not. Even the "production OD grinders" required indicating mics for final size. That was the late 1970's and my comments/opinions are derived from what the professionals did. Perhaps things have changed.

CMM's- Fast forward into the 80's and I'm doing tool and die work for in-house stamping. It was SOP to take a first part into inspection for verification if you changed something before a run. They had a large Cordax CMM that often showed a reading different from what you measured in the tool room with your own mics/indicators. When I queried the inspector he told me that the mfg. rep stated that it was only accurate to around +/- .001 and should not be trusted closer than that. I'm fairly sure that things have improved since then and I see published accuracy/resolution/repeatability numbers well below .0005 but I often wonder just how much trust can be put in those numbers. Like many other things, more money produces better results but I moved into engineering and haven't had access to a late model CMM in good working order in may years. Others have posted using more than one device for dimension verification and I'd agree.

Yes it is nice to see other metal workers on this board. Hope everybody is doing well. Ganbatte.

 

[RoninB4]

Holding 0.0001" is no joke and it only gets more ridiculous when they give you a 50 millionth tolerance (0.00005").

I liked a good challenge now and then but tried to avoid a steady diet of work that close, too nerve wracking chasing .0001 for square and parallel all day every day. Maybe I have a short attention span.