In reality, a surface plate in the manufacturing shop is not always the highest caliber table. In most manufacturing even climate controlled areas .0002 is more than acceptable. You will be searching for bigger defects that a couple of tenths. As a quality engineer, I would be looking at a root cause analysis! Validate your machine and machine tools are capable of holding those types of tolerance . Keep up the great work, great video!
Just waking up here, idly drinking coffee and already learned a few things! I like this level of instruction, little bits of good practice... and especially a topic like this, that might help you recover when things aren't going quite right. You might be able to make a diagnosis and keep going.
I look forward to every vid you put out they are so useful.. Thank you very much. They don't help my OCD lol. With having everything bang on the money thow.. Take care bud lee
I don't quite understand the role of the gauge block in this process. I understand zeroing out the dial indicator, but why do you add a gage block underneath? It would add an additional specific height to the plate, but I can't figure out why this is necessary. Thanks for all the videos!
saw an article about inspecting lathe beds to tenths with a laser, it will work for surface plates too. Blast a laser beam across its surface and observe the scatter on the wall behind.
We had ours resurfaced once or twice in my years at the shop. Let me tell you the guy that did it had big arms! Its a lot of work getting .0005 back to flat! They were 4'x4' and 4'x5' and several 2'x2' I feel you got an fairly accurate idea of the condition of you plate. Always try to use the 'best' section for your job.
All of your videos are great IMO. I am just a hobby hack but still enjoy the professional level videos such as the ones you put up. It takes a bit of nerve to put up one like this that shows the non professional way to determine flatness because there are always folks looking to put down and/or discredit others. Let's face it, in some instances, this check is good enough right? I'm wondering if it might be possible to check a plate with a wire? Maybe a strong and thin wire stretched tight across the plate BUT not actually touching the edges of the granite? Instead it would be supported at the edges with round spacer having 1/4 of the circle removed so they fit over the edges? Then a gauge block and feeler gauge used between the plate and the wire?? Just curious and figured you would know the feasibility. Thanks
Any way for hobbyists to do this with some other equipment? I have thought about taking something with two points, one dial indicator in the middle, set zero at some arbitrary point and then go around, repeat with the two outer points at different distances. That should work even with cheap $5 chinese dial indicators and a scrap piece of steel, shouldn't it?
I think I've seen pictures of this apparatus! Calibration sounds like a problem, you need something straight to set this with. Just from what we learned today, it's looking like you might zero your indicator along one edge of the plate, and then put this bridge across the center in different ways (crosswise, diagonally)? Ah, your HFT indicator probably only reads in thousandths! It might be that for the hobbyist, the granite plate and the cheap Chinese set of gage blocks are better than anything else we have, and we just have to BELIEVE in them! That's what I do! The kind of errors he's talking about here are hard to measure at all with 'hobbyist" level equipment. There's a famous and rare book "Foundations of Mechanical Accuracy" by Moore, I got one through Interlibrary Loan, but it's also online as a big .pdf file. If you struggle with this a little, you might be on your way to becoming a tool maker or gage maker, and maybe the last one!
I recently discovered your channel, it has been very helpful in refining my machining skills. I am a bit confused by what you mean when you say "tenths" . From context is sounds like you are talking about a unit smaller than a thousandth(0.001"), but a tenth [of an inch] is (0.010"). So is "tenth" short hand for a ten-thousandth(0.0001")? If so, is that common industry speak?
You know, you can get better than 1/10th of an inch with a hacksaw, so it's not very interesting to talk about! People were complaining about this bit of confusion terminology a hundred years ago, but it hasn't gone away. I think in this context, if you were talking about 1/10" you'd have to say "Point one inch" or "a tenth of an inch"- but he might be talking about .0001" all day long, so that gets the shorthand..
I see the reason for the gage block. There is less chance of losing your reference than just dragging the indicator along an unknown surface height at a steeper angle.
No. It would help maintain measurement accuracy because the needle needs to be parallel to the surface to prevent sine errors, but in this task the actual measurements don't matter too much - they are only being used to get a sense of plate accuracy, not a calibration. An error of 20% would be terrible if calibrating, but isn't really a problem here. A reading of 1.6 tenths vs 2 tenths isn't important - we're looking for many tenths or even thousandths as an indication of a problem. Also, if you know the needle angle you can compensate with trigonometry anyway. Knowing that a needle at 45 degrees will read 0.707 of the actual movement allows you to multiply by 1.41 to correct. I bet most would struggle to set their needles that accurately though, but at least you can reduce the error. The reason for the gage block here is that although surface plates are flat, they are not necessarily smooth. High spots are removed when lapping flat, but divots still exist. In normal use they don't matter - the feet of the gage mount, and the base of the part being measured are all large enough that they ride the flattened high spots and bridge those holes. The needle of the indicator is not large enough to ensure that, and could as a result drop into a hole and under-read. Here, the gage block acts like those feet would, and being incredibly smooth and uniform in thickness, presents a consistent surface for the indicator to measure.
I have one of the little imported granite plates, and am trying to get someone to measure it with a co-ordinate measuring machine, just make a hundred touches on the surface or something, try to make a little map? I work in a big Engineering College so you'd think it would be easy. I don't think my school specializes in metrology, and it may be that they OWN this equipment but it's just sitting. This would not be the best method, but seems like it would be quick and easy. A good surface plate (which I might not have) may be near the limit of resolution of the CMM.
There are algorithms that solve the real surface feature by two linear measurments of unknown quality but my mind boggles even right when I see a 0.01mm difference of reading flat. To make matters worse mine shows a compound error of 0.06mm...
Thank u sir. Sir make a vedio about how to check machine bed with different method is it parallel or not. And how we do masuring instruments caleberation.
There is a synthetic diamond manufacturing company moving into my lil town soon. I'll stop by and get a quote for a 18 x 24 diamond crystal plate that is perfectly flat. Their answer should be interesting. I know they will say no, we cant make that. What I really want to know is this, if you could make one, how much would it cost?
paper towels leave fine fragments, better to use a shop cloth ... if you have a clean one .. they are also abrasive, you could strop a straight razor on paper towels.
How about microfiber cloths? They are cheap now, after being washed a few times seem very clean and seem to leave a brilliant clean surface. I have seen optical gear cleaned with these, it gave me the idea. It might be that they need to be kept covered or in a box, and changed frequently. They're not very absorbent, but seem good at picking up fingerprints, maybe with a little spritz of alcohol first? I'm going to try ammonia first thing Monday morning!
