What's new
What's new

hinging parts

dian

Titanium
Joined
Feb 22, 2010
Location
ch
how much does a part have to be off to hinge wrong, lets say on an aa plate? e.g. a part 1x1", 5" long.
 
exactly. if i take a part and measure it 2.5 mu high in the middle it will hinge dependent on where i put it on the plate. 2.5 mu and the tolerance of the plate almost add up to your 0.0003". i have, however, a mysterious parallel, that i measured and remeasured with several instruments and different setups. its completely flat, my best guess is that its 0.5 mu out of parallel but flat. it stubbornly hinges on the ends wherever i put it on the plate. i stoned the edges, oiled it up, put a very light coating of blue on it, no difference.

any thoughts? im using a din 876 00 granite plate (total deviation 3 mu).
 
My friend and student Stefan G has a great you tube show on scraping and hinging. It is exactly the way I teach it in my classes. I use 30% and not 33.3%. look for hinging around 20 minutes. Depending on the surface plate accuracy, thickness of the bluing, hand heat, shop temp, etc.. If all was right I say you can hold .00005" per foot. I do it all the time when scraping straight-edges.
YouTube
 
the offending parallel blues up like this:
 

Attachments

  • 0 157.jpg
    0 157.jpg
    93.7 KB · Views: 175
Are we sure that it doesn't have any bur at the ends? They could be very hard to see, but tall enough to mess with the hinging and, at the same time, small enough to allow the bluing of the rest.
I like Richard's suggestion of robbing against a portion of the plate not coated with ink.

Paolo
 
as mentioned i stoned the edges, i also rubbed it dry on the plate. if i take the parallel, lightly blued as on picture and rub it on the plate, the plate blues up in a uniform way. i did i five times (about 5 seconds rubbing), cleaning plate in between an only then the blue was gone on the edges and a slight haze remained in the middle. so how much is that out? 0.25 micron?

ironsmith, you say it works to that level reliably. but how so? you clearly get into the accuracy of the plate. look at this:
 
this is 2-3 mu out. it hinges either corrrectly or in the middle, depending where i put it
 

Attachments

  • 0 144.jpg
    0 144.jpg
    89.8 KB · Views: 102
from the looks of the part it looks like you have sanded it on emery cloth that was laid on a surface plate as the lines under the blue are not straight from a grinder or a scraped surface.. ?? If your doing that, the part will get high in the middle as you can't control the wiggle or rock (like a rocking chair) of the part when the end goes off the cloth / paper when you hand push and pull it back and forth.
 
tg, thats what im saying. you have to take the variation of the plate into account. (i think i have better plates then most, din 00 more or less corresponds to aa.)

the above part has been blanchard ground and yes, i gave it a few strokes on emery cloth to make sure there were no burs before bluing.

in the meantime (and this might save others some headaches) i found out what is going on with the parallel. its so simple when you know it: the short side is not parallel, it rises up to 4 mu over the 8 mm. so any measurements along the 200 mm are meaningless, i would have to check the high edge somehow and even that would not very well predict how it hinges.

this really opened my eyes as to the use of parallels. if you want precision on the mill you have to put something flat on them and check that surface first. its also surprising what happens with this surface when the vice is tightened despite being tapped down. now i appreciate my hydraulic vices even more.
 
Is your surface plate from a reputable maker? Did you purchase it new? If not new, have you ever had it inspected by someone with the proper equipment? It shouldn't be hinging appreciably differently on different parts of the plate if it's an AA equivalent.
 
I work in a calibration lab, and I've seen hinging work all the way down to 10-20 millionths reliably, or .25-.5 micron in metric.

I think that says it all. Hinging a part tells you something, middle/edge points of contact, but not by how much. When done with no blue the test becomes super sensitive.
You could worlds flattest reference and finest scraped straight edge. With the smallest heat input in a less than ideal spot, that SE could go from middle contact to edge contact on a dry surface and still print well either way. I think too much is made of the test sometimes, especially by the new guys.
 
I will add that in a calibration setting, hinging is not a final say on any calibration process, just a trick to help you determine the quickest method to follow up with. The plates used are of laboratory grade and we have the printouts showing us how the plate looks in topography. Checked yearly. Yes the error of the plate is 50% of the reading but is so well known in this case it can be negated or at least accurately factored in.
 
dian;3521651i have said:
the short side is not parallel, it rises up to 4 mu over the 8 mm. so any measurements along the 200 mm are meaningless, i would have to check the high edge somehow and even that would not very well predict how it hinges.
I don't understand how you could measure 0,5µm (0,2 tenths) when your parallel is off 4µm (1,6 tenths) ?
Be careful when measuring µm dimensions 100mm -> 1° -> 1µm (3,9" -> 1,8F -> 0,4 tenths)
When you hold this small parallel 1 Minute in your hand (in the middle) yout can be shure that it is high in the middle

greatings, Franz
from the Lüftinger Brothers
 
I don't understand how you could measure 0,5µm (0,2 tenths) when your parallel is off 4µm (1,6 tenths) ?
Be careful when measuring µm dimensions 100mm -> 1° -> 1µm (3,9" -> 1,8F -> 0,4 tenths)
When you hold this small parallel 1 Minute in your hand (in the middle) yout can be shure that it is high in the middle

greatings, Franz
from the Lüftinger Brothers

very simple: i took readings in the middle, 4 mm from the edge. there it was flat. and no, i dont hug the parts, when working on the plate.

franz, you might know how it is over here. most parallels seem to conform to it5 (i rarely see an it3). meaning the working side (precision defined by height) is less precise than the flat side (precision defined by thickness)? because the height is larger that the thickness? and then the precision across the working side (my 8 mm) is defined by the lenght? it doesn make too much sense, but this is what it seems to be like.
 
I was thinking have you checked the surface on the short side. You can lay a gage block on it and check the hinge. It looks high in the middle to me. I'm curious what does the other side look like? A few more pictures from afar showing us your plate, height gage, how your set up is. No guessing then.
 
not sure what we are talking about now. the block with the hole was just an example of a part i knew hinged inconsistently. 2 mu is apparently enough to put deviations of the plate into play, which makes sense, as total deviation is 3 mu (i would have to look up what is allowed locally).

my "offending parallel" ( 150x40x8 mm) was a measurement error, i measured just the middle line along the 150 mm. i should have mapped out the whole 8x150 mm surface. as mentioned in the previous post, it was a surprise there could be a larger deviation across the 8 mm that across the 150 mm. but the whole set is like that. i dont remember what the other 8x150 mm side looks like, cant blue it up because not sure which one it was. (i just marked the two best parallels with a 0.5 and 2 mu deviation.)

hinging was done on the large plate and measuring on the stand:
 

Attachments

  • 0 123.jpg
    0 123.jpg
    88.9 KB · Views: 87
  • 0 138.jpg
    0 138.jpg
    95.3 KB · Views: 60








 
Back
Top