What's new
What's new

Scraping in a precision master level

dcsipo

Diamond
Joined
Oct 13, 2014
Location
Baldwin, MD/USA
As promised, i am starting a new thread, I took my levels to the surface plate, just to realize that i have a lot of cleaning up to do in the shop before i can start this project, I was thinking if i am going to to this it should be a twofer. I have two Starrett levels: a new to me, older 199 and an even older 98-12. As all good stories start, here is chapter 1.

The only thing i did today is a bit of discovery. Stoned both levels lightly before the start.

Discovery #1: was the mess in the shop, but immediately after that
Discovery #2: the plate not being level. Since it is sitting on a concrete ledge covered with some framing and flooring it is not bedrock, but not a bouncy floor either. I proceeded to level out the granite which led to
Discovery #3: the level is not calibrated. So i calibrated the 199 using the usual process. Got the plate level with the level in one direction, blocked in the level in XY, flipped it 180 adjust to split the difference, flip verify the offset is the same on both side, re-level the plate in X&Y, checked on 180 flip. Plate is level now and the level is calibrated.
Me still being a slob, and not in the mood to clean up the other project, i just blued up a small part of the plate and blued up the levels. that led to
Discovery #4: As you can see IMG_2234.jpg the spotting is a bit strange pattern for a level. Not exactly what i expected. The edges touch the middle is hollow, also touches in the center, still hinges close to the outer edges of the ends, good news it makes it easier to scrape and spot, the level bottom is fairly flat. This spotting indicates that someone surface ground it or had a very good sense how to lap something like this. At this point i had an ohwhatthehell moment and blued up the 98-12. As luck would have it i made
Discovery #5: The 98-12 is actually shaped like a level should be, since it cannot stand on its head i took a picture with the 199 supporting it IMG_2235.jpg. There is wear on the corners and the bottom is scratched, but it is a good start. So tomorrow after my bike ride, i will get the gumption to clean up the mess and get into some scraping. But i had enough stuff out to try one more thing, i qualified a spot on the surface plate to be level, blocked in the 199 then used a couple of 123 blocks to mark the perpendicular center of the level, turned the level 90 degrees keeping the vial center approximately in the same spot, the cross level looked like it was dead on, the bubble smack in between the two lines as I looked at it with a 2x magnifying glass. Then i located some of the 0.0005" mylar shims i got from Nathaniel. (I could not find anything thinner. )Even sticking one shim under either edge of the 199 shoved noticeable change on the cross level. and that was
Discovery #6 for today, if you off more than a tenth or two the little cross level will show it, i was not expecting it to be this sensitive. Extrapolating it a bit further the cross level is about 0.003"/12" or better. Once i finished i will measure it in a more sensitive setup, W\will put my CI surface plate on the granite and shim that with thin stock.

After all this i gave up and came in from the shop.

Hopefully i will be getting to the next chapter tomorrow.

dee
;-D
 
Starrett scrapes their 199 levels with a deliberate 0.0005" concave in the face to ensure end bearing. Most precision level makers do - or did: Lufkin, Pratt & Whitney, etc

The idea is to bridge short work with equal overhang and on long work the level rest on the ends. This is generations old standard practice that I chose not to follow. I scraped all my levels flat, no hollow. I use them for short scraped straight edge. Handy: a straight edge AND a level.

Scrape your precision levels flat if you wish but be aware of the traditional practice and the reason it's lasted so long.
 
Starrett scrapes their 199 levels with a deliberate 0.0005" concave in the face to ensure end bearing. Most precision level makers do - or did: Lufkin, Pratt & Whitney, etc

The idea is to bridge short work with equal overhang and on long work the level rest on the ends. This is generations old standard practice that I chose not to follow. I scraped all my levels flat, no hollow. I use them for short scraped straight edge. Handy: a straight edge AND a level.

Scrape your precision levels flat if you wish but be aware of the traditional practice and the reason it's lasted so long.

.
wyler sells levels where center 1/3 of level is relieved or cut back like 0.100" to make sure level sits on ends and does not rock. a level that rocks giving different readings is like a dog chasing its tail.
.
and better levels are not flat but got a Vee base or involute base for use on cylindrical rollers. on a paper or film machine 99.99999% of leveling is done on cylindrical rollers.
and a very high precision level can easily pick up the sag of a roller from gravity and you need the cross level vial to make sure level is horizontal or at top dead center of roller.
.
you might want to make a separate scraped tool for checking flatness than a level.
 
level i made. low profile, Vee base, can adjust vial up and down and sideways
.
sideways vial adjustment and the cross level vile needed for cylindrical rollers
.
deliberately open design to fit in tight spaces and aluminum used as it does not rust. actually always works very good. i made in 2003
.
i have milled level bases and if i felt it rocked i would relieve the center 1/3 so it sits on the ends. but never had too. properly milled always worked good enough
.
2nd picture i added a oring "spring" and knob to calibrate in seconds. and notice like most Swiss designed level vial adjusts sideways too. not just up and down
 

Attachments

  • LevelLowprofile.jpg
    LevelLowprofile.jpg
    72.5 KB · Views: 347
  • level2.jpg
    level2.jpg
    83.3 KB · Views: 348
level with bottom sides and top with milled Vee. can use on top, bottom, and sides of cylinders
.
i never scraped Vees as i never needed too. if it rocked (it doesnt) i would have cut back or relieved center 1/3 part
.
made in 2002, vial from a Starrett 98 level and using plastic tipped set screws. it always works good. biggest thing is to not pick up the roller. the rollers having bearing clearances on each end. its possible to literally pick up the roller easily .003" per 72" length of roller. or effect the sag of the roller from gravity
.
i am a former millwright who often made custom tools depending on what i was doing and machines being worked on. notice Starrett 98 level with brass quick calibration knob and the florescent / glow in dark paint behind vial you can see 10x easier
 

Attachments

  • leveluniversal.JPG
    leveluniversal.JPG
    68.7 KB · Views: 289
  • Level3.JPG
    Level3.JPG
    47.9 KB · Views: 232
  • Level4.jpg
    Level4.jpg
    78.4 KB · Views: 354
  • level5.JPG
    level5.JPG
    75.3 KB · Views: 271
Frame Level with extra medium sensitive level vial added so not always reading off scale
.
once i added the less sensitive level vial i used it 10x more often
.
this old millwright added a quick easy fast calibration knob of course.
.
i have for checking vertical ball screws to check if really vertical or plumb
 

Attachments

  • LevelFrame.JPG
    LevelFrame.JPG
    58.4 KB · Views: 272
I don't know what is all this about but if you need a flat level just fly cut it and then take it to somebody who's got a lapping machine. You'll never scratch ( sorry - scrape ) metal to the accuracy of a lapping machine. Once it's lapped, scratch it with small strokes to make it look "professional". A newish grinder will give the same results more or less.
 
Starrett scrapes their 199 levels with a deliberate 0.0005" concave in the face to ensure end bearing. Most precision level makers do - or did: Lufkin, Pratt & Whitney, etc

The idea is to bridge short work with equal overhang and on long work the level rest on the ends. This is generations old standard practice that I chose not to follow. I scraped all my levels flat, no hollow. I use them for short scraped straight edge. Handy: a straight edge AND a level.

Scrape your precision levels flat if you wish but be aware of the traditional practice and the reason it's lasted so long.

I agree with Forrest. I have scraped dozens of Starrett Levels and tested new ones. The factory grinds them and and then square cuts them to break up the surface. They blue up about 30% from each end with the middle relieved like he said .0005".

Dee is an excellent scraper and will have no problem doing it. Also his friend Paolo could easily help him. Both were students at the Tuckaho Class. Just be sure to press down on each end when done to assure the middle has been relieved enough. Pushing it down while watching the bubble. The bubble should not move.
 
scraping a level may not be necessary but it is done on many precision instruments. Also it allows you to make changes to instruments even if you dont have more expensive equipment. Third point there is a great deal of self satisfaction with doing a job yourself, even if you have a bit of help.

I hope you have continued success in your efforts. Some of the ideas represented here sound like a good idea. I think you may be overthinking it a bit, perhaps after the Turkey settles out of your system you will find you can concentrate better, or at least perhaps see things a bit differently. Thanks for sharing.

Charles
 
.
wyler sells levels where center 1/3 of level is relieved or cut back like 0.100" to make sure level sits on ends and does not rock. a level that rocks giving different readings is like a dog chasing its tail.
.
and better levels are not flat but got a Vee base or involute base for use on cylindrical rollers. on a paper or film machine 99.99999% of leveling is done on cylindrical rollers.
and a very high precision level can easily pick up the sag of a roller from gravity and you need the cross level vial to make sure level is horizontal or at top dead center of roller.
.
you might want to make a separate scraped tool for checking flatness than a level.

Tom your right too. There are several styles of levels. I am sure you have seen a .0005" Starrett level too, haven't you?
they are flat. http://www.practicalmachinist.com/v...t-precision-levels-starrett-exact-level-2.jpg
 
scraping a level may not be necessary but it is done on many precision instruments.

That is indeed very true. But I found a precision level works noticeably better when ground/lapped and not scraped. Neither my Zeiss level or clinometer are scraped - they thought it would interfere with the measurement.Of course, nobody knows everything and Zeiss might've been wrong.
 
Lapping might be better if you have access to the facility but scraping is accessible to anyone.

Could you even lap a box square on four sides and keep it square to working tolerances?
 
got a little time in today. I felt i was a bit out of practice so started with the less critical 98-12 level, it has a channel in the bottom which breaks the surface in two narrow strips. Cleaned up the shop a bit then started on it.

IMG_2245.jpg

got a fresh surface on it with a 90 mm blade, spread some blue...

IMG_2247.jpg

spotted it...

IMG_2248.jpg One end of the spotting is washed out in the light, but it is about the same as the other end.

then scraped it till i got it to hinge right, then switched to a 60 mm blade and got it this far

IMG_2263.jpg

still has some areas that have only very light blue and some places it is darker then i like it...i quit for the night..cleaned it up, the pattern is starting to develop. It is hard to scrape the narrow strips evenly.

IMG_2265.jpg

I may relieve the center or may not, have not decided yet.

dee
;-D
 
Looks good, but remember use a narrow blade about the width of the surface or narrower and to shorten the stroke a little at a time. On a narrow surface like that I would think a 15/150 - 5/8 wide x 6" long and a 1/8 to 3/16 stroke length and a 60r blade. The chart I gave you says 90 r for roughing. But that is only if the part is out more then .001" I am assuming yours was not. As far as leaving the center low. If your going to use it as a straight edge or parallel to measure alignment while scraping leave it flat, if your going to use it to level machines then relieve the middle so when you "seat" it wears better.

In both cases when it is finished, press down on both ends of level to "prove" it isn't high in the middle and doesn't rock like granny's rocking chair. On a .005" per division you may want to use a magnifying glass to see if the bubble moves. Fine Job Dee! I am very proud of your work and willing to share it. Rich
 
More scraping tonight, i think i will not relieve the 98-12 just leave it flat.


here are some pictures of today's results

IMG_2266.jpgIMG_2267.jpgIMG_2268.jpgIMG_2269.jpgIMG_2270.jpg


Will start on the 199 Starrett in the next few days.


BTW it a bitch to take photos of this.

dee
;-D
 
Last edited:
That is indeed very true. But I found a precision level works noticeably better when ground/lapped and not scraped. Neither my Zeiss level or clinometer are scraped - they thought it would interfere with the measurement.Of course, nobody knows everything and Zeiss might've been wrong.

I've never owned, used, or even seen a level that has been scraped on the base. That's not to say you shouldn't do it if it pleases you.

Regards Tyrone.
 








 
Back
Top