Methods of measuring overhang height using gauge blocks?

I often prefer to measure dimensions as a height above a surface plate. For example, if I have a block I will measure its top surface's height above the plate. I don't really trust my vernier height gauge to better than a thousandth or so. So, if I want to measure to say +/- 0.0002" it can get pretty tedious because I make a preliminary measurement using the height gauge, then make a gauge block stack to that height, then check the stack against the work piece, then adjust the stack and repeat the process until I have it within 0.0002" consistently. To do the comparison I use a test indicator mounted on a surface plate stand.

It's especially problematic to do this when I need to measure the height to an overhang. To do this I invert the arm of the height gauge so it touches the underside of the overhang and compare the arm to the gauge block stack. The problem here is that I am not actually comparing the block stack to the overhang, but to the arm, so I am worried about the additional error that introduces.

Does anybody have any suggestions on technique using this equipment (gauge blocks and an ordinary precision vernier height gauge)?

Unless I'm missing something really obvious, it seems to me you are doing this the hard way just because you're used to it. It also seems that you don't have the right tools to measure +/-.0002" reliably. With basic tools, I would measure your two examples with outside mics and a depth mic.

Matt@RFR said:

Unless I'm missing something really obvious, it seems to me you are doing this the hard way just because you're used to it. It also seems that you don't have the right tools to measure +/-.0002" reliably. With basic tools, I would measure your two examples with outside mics and a depth mic.

Click to expand...

If you are talking about holding a micrometer in the hand, I don't really consider that remotely comparable to a measurement done on a surface plate. The micrometer will tilt. Also, some parts have shapes that are difficult to just hold in your hand and bracket reliably with mic anvils. You can prove all this by just taking a block and asking 5 different guys to measure it with a mic, and you will get 5 different answers. I would never use a mic to measure anything tighter than 0.001". If you read a textbook on metrology or guidelines from a maker like Starrett, you will see that they completely agree with this because they say that the so-called "designated precision" of a vernier micrometer is 0.001".

Any normal inner micrometer cannot measure overhang heights less than an 1" and even if you could you would have tilt problems. Also, most normal micrometers have a best case accuracy around +/- 0.0003" so they are not even remotely comparable to gauge blocks which have an error of 0.25 micrometers (0.00001"). In other words a mic has at least 30 times greater error than a mic. In practice I find mics tend to be even worse than that because of various mechanical and feel-related issues.

jscpm said:

I often prefer to measure dimensions as a height above a surface plate. For example, if I have a block I will measure its top surface's height above the plate. I don't really trust my vernier height gauge to better than a thousandth or so. So, if I want to measure to say +/- 0.0002" it can get pretty tedious because I make a preliminary measurement using the height gauge, then make a gauge block stack to that height, then check the stack against the work piece, then adjust the stack and repeat the process until I have it within 0.0002" consistently. To do the comparison I use a test indicator mounted on a surface plate stand.

It's especially problematic to do this when I need to measure the height to an overhang. To do this I invert the arm of the height gauge so it touches the underside of the overhang and compare the arm to the gauge block stack. The problem here is that I am not actually comparing the block stack to the overhang, but to the arm, so I am worried about the additional error that introduces.

Does anybody have any suggestions on technique using this equipment (gauge blocks and an ordinary precision vernier height gauge)?

Click to expand...

.
use different height gage like a digi chek
.
JTS Machinery & Supply Co.
.
has steps you can directly set test indicator too saving the use of gage block

After sitting down and thinking about this for a while it occurred to me that I can measure the overhang directly if I add an extra block to the stack. It's obvious once you think about it. Just wring an extra block on the stack and rotate it 90-degrees to make a mini overhang. Then a test indicator can be used to directly check the overhang on the stack against the work piece.

I guess I never thought of that before because I don't measure overhangs too often.

DMF_TomB said:

.
use different height gage like a digi chek
.
JTS Machinery & Supply Co.
.
has steps you can directly set test indicator too saving the use of gage block

Click to expand...

Yes, I know there are $5000 height gauges that go to 0.0001". I specifically said I want techniques that work with gauge blocks and an ORDINARY precision vernier height gauge which is what I have.

gage block accessories is used for many things
.
https://www.google.com/search?q=gag...ChMIwpGIydKayQIVARUeCh291wbU&biw=1000&bih=601
.
i often put one or more ends on to use as height gage, measure a OD gap and with round ends you can use to measure a bore.

jscpm said:

You can prove all this by just taking a block and asking 5 different guys to measure it with a mic, and you will get 5 different answers.

Click to expand...

Sure, then hand the same setup to 5 WELL TRAINED people and get 5 identical readings. The stuff we check here with the methods I mentioned get verified with CMM's and vision systems, and we're almost always spot on.

If you're talking about production, then yeah, setup gauge block stacks and indicators on the surface plate. Doin' 5 parts like that and you're wasting your time. At least that's how it is in my shop.

Adjustable parallels. Set the adjustable parallel to the overhung recess, take it out and measure its height. Just sayin'

jscpm said:

After sitting down and thinking about this for a while it occurred to me that I can measure the overhang directly if I add an extra block to the stack. It's obvious once you think about it. Just wring an extra block on the stack and rotate it 90-degrees to make a mini overhang. Then a test indicator can be used to directly check the overhang on the stack against the work piece.

I guess I never thought of that before because I don't measure overhangs too often.

Click to expand...

Never needed to measure a overhang to tenths, but what I've done is to stack the hight I'm looking for then wring another block with an overhang and zero dial test indicator to zero to this hight under the overhanging block.

Then sliding this under the over hang will indicate the plus or minus from the stack height.

Instead of using the guage blocks and vernier you could get hold of a height micrometer ,they seem to go pretty cheap

Fowler Trimos Height Micrometer

jscpm said:

After sitting down and thinking about this for a while it occurred to me that I can measure the overhang directly if I add an extra block to the stack. It's obvious once you think about it. Just wring an extra block on the stack and rotate it 90-degrees to make a mini overhang. Then a test indicator can be used to directly check the overhang on the stack against the work piece.

I guess I never thought of that before because I don't measure overhangs too often.

Click to expand...

That's one of those dope slap moments, yes? I get them often so don't feel alone.
Bob

Matt@RFR said:

Sure, then hand the same setup to 5 WELL TRAINED people and get 5 identical readings. .....

Click to expand...

What is real fun is to take 10 .7500 gauge blocks. Lap them down to .0002 to .0020 under leaving the size marked on them.
Better yet if you have some oversize in the mix which means you have to chrome up one side and then lap.
They need to all look the same. Basically trying to fool your people.
Give them randomly mixed to 5 well trained guys to measure with any manual measuring device Do this 3 times across a couple of weeks.
Map the results, plot the curves. Very time consuming but an eyeopener for sure.

You will most likely swear on your Grandmother's grave that you can do better than the results show and ask for another try .
I know I did, a mic in my hand was the most natural feeling in the world and a tenth or two was certainly in my ability if you cleaned everything and were careful.
I had measured thousands if not tens or maybe hundreds of thousands of parts with tolerances of +/-.001 being the loose guys.
Gets even better with round gauge pins.

It's like being told Santa-Claus might not be real. Your brain just does not want to go there.
Bob

Just hope some real trickster does not slip one of those blocks in a working set..

You still got that random set Bob? I'd love to rent it from you! That would be a fun experiment, but I don't have the equipment or knowledge to lap.

Slipping one in, that would so much suck.
But they are part of The R&R sets and do have microscopic marks on them so that we can sort them for the random handout.
They are also a set used for this purpose so if one was missing we'd go into panic mode fairly quickly.

Matt, no way these are leaving. way too much money getting them NIST certified and I no longer have the lapper. Same with my holder taper gold masters.
Near and dear to my heart and locked up for careful use only.
You can do much the same by grinding down some cheap ones. You only get under this way and it gives a clue out but still works. You need a decent variation and one or two flyers.
Random drill rod also works but you may need to mark where to measure so everyone is on the same page as it might not be round.
Actual scrap parts are the best as this is what you make and measure.
They need to look identical and in a different order so the guy measuring does not "remember" what he wrote down three days earlier.
We do blocks, rod and actual inserts along with angle checks where we toss in some curve balls where the angle changes depending on where/what you measure.
It's part of the training/ operator certification program.
Fail this outside the expected norm and you don't get the raise.

The big thing is that the person measuring does not know the actual size or better yet is intentionally misled by the markings or their knowledge of that part.
You can't use a known good gauge block set. All the readings will come out good.

I'm a big believer in that if you can measure it, you can make it, no matter the number of zeros. I'm kind of a nut here.
Bob

CarbideBob said:

What is real fun is to take 10 .7500 gauge blocks. Lap them down to .0002 to .0020 under leaving the size marked on them.
Better yet if you have some oversize in the mix which means you have to chrome up one side and then lap.
They need to all look the same. Basically trying to fool your people.
Give them randomly mixed to 5 well trained guys to measure with any manual measuring device Do this 3 times across a couple of weeks.
Map the results, plot the curves. Very time consuming but an eyeopener for sure.

You will most likely swear on your Grandmother's grave that you can do better than the results show and ask for another try .
I know I did, a mic in my hand was the most natural feeling in the world and a tenth or two was certainly in my ability if you cleaned everything and were careful.
I had measured thousands if not tens or maybe hundreds of thousands of parts with tolerances of +/-.001 being the loose guys.
Gets even better with round gauge pins.

It's like being told Santa-Claus might not be real. Your brain just does not want to go there.
Bob

Click to expand...

You have a lot more optimism than I do, or more likely, you work with more competent jokers. It's pretty rare that I have known or encountered a machinist I would have any confidence of making a repeatable sub-thousandth measurement.

jscpm said:

You have a lot more optimism than I do, or more likely, you work with more competent jokers. It's pretty rare that I have known or encountered a machinist I would have any confidence of making a repeatable sub-thousandth measurement.

Click to expand...

Are you saying there is no Santa-Claus?
This tenths argument comes up here often. People don't want to hear this. They want to believe, they feel it in their bones.
They just know they can do it. You will not convince them otherwise.
Bob

Whynot -.0250

Heavey Metal said:

Whynot -.0250

Click to expand...

Is this to me and testing blocks?
Because you want to bias the guy doing the measuring . If -.025 it is obvious. (maybe not for some.)
I guess this a good check for standard mics but we can't usually make the one turn error and it's not what we are after here and would totally screw the math for the distribution curves.
Blowing that is oh my gawd, silly oops and a flyer that you would simply toss out.
Sure sometimes we may have all done this mistake but it way outside of measuring accuracy and it's kick me upside the head stuff.

You are looking for the worst you get out of a thousand measurements not the best. You are trying to find the maximum error so you can plan for it.
If a person is making 1.000 shafts or holes his/her measurements get biased towards the print nominal. It's just human nature.
If you take a part with some weird number like .4463 and have people check it you will get different numbers than if you give them the print first, show them nominal and then they measure.
Many would say it should not happen but in real life it simply does. That is the way our brains work, like it or not.
You want to see this bias. You play sneaky tricks to actually see and know it.

It's a subconscious thing and we all think we don't do it but we all do.
Ever measure a part and it's out, check again at it's in. Two checks, neither one more valid than the other.
Yet most will accept the good one, assume the out was a bad check. Better yet when there is a third check and it is borderline.
One out, one in, one maybe?? Good part bin or bad part bin?
I know where I put it. What do you do?
Bearing bore is maybe on the high end, shaft is maybe on the low side. fuss a bit with measuring..........
Yea, tell me you have not been there and I'll say you don't have enough chips in your shoes.

I darn well know where tenths and micrometers lead too.
The land where Angels fear to tread.
Bob