What would be considered "typical" tolerance when scraping parallel and perpendicular?

[marka12161]

I'm scraping a 3" x 3.5" x 6" cast iron V block as a practice project. I've got the bottom flat and am about to move on to the top flats then the sides. What tolerances should i be shooting for? I'm thinking something like parallel within .0001 over the entire surface and perpendicular to .0001 over 3 inches. Do these seem like reasonable targets?

 

[Richard King]

Super precision scraping is .00005 / 12" or that's what I was taught on inspection equipment. One needs temperature control room and you need to wear cotton gloves so you don't transfer body heat to the parts. The Moore book has a great chapter on Temp. Control and tolerances.

 

[sfriedberg]

0.0002" in 6" is a fairly common spec flat/square for "good" ground surfaces. I just looked at a couple of things on the Suburban Tool website, and they spec a few things to 0.0003" (distance unspecified, but I'd assume 6") for flat/square, while their cube parallels are spec'd to 0.0002" in 8". Some items are spec'd even tighter. Myself, I would shoot for 0.0002" in 6".

 

[marka12161]

0.0002" in 6" is a fairly common spec flat/square for "good" ground surfaces. I just looked at a couple of things on the Suburban Tool website, and they spec a few things to 0.0003" (distance unspecified, but I'd assume 6") for flat/square, while their cube parallels are spec'd to 0.0002" in 8". Some items are spec'd even tighter. Myself, I would shoot for 0.0002" in 6".

Thanks. Makes sense. I guess with a tenths indicator, you can only reliably measure to .00005 over the height of whatever part you are trying to square up.

 

[lucky7]

I found scraping a pair of v blocks a great project. Had to ask for advice here on how to get the v centered and at 45deg. In my non Moore quality room, I’m not hitting half a tenth. Nor can I realiably measure to that fussiness with my grade A plate and my gack. What condition is your plate? How stiff is your method of holding your tenths indicator and is it 3 pointing on plate or is there rock? Many other questions.

2 tenths over 12” is something I can reliably hit (I’m an amateur). I have done 60 ppi on two smaller projects and never want to do that again. I’m sure with practice you can do the same or better. Just please listen to people smarter than me and don’t chase better than needed accuracy.

 

[MCritchley]

It’s surprising how easy it is to get something with in a few tenths via scraping. After your few tenths excitement cools down you can move onto 1 or two tenths accuracy. Then after you’ve gotten that close go for zero, it may take 50 passes but it can be done. By zero I mean no more than a tenth.

Do you have a good indicator, and is the bottom of your surface gage scraped? First your inspection tooling and surface plate needs to be good or you’ll be chasing your tail. Lastly get a good parallel that you can lay on the scraped surface to check the slope without having to worry about indicating on the high points.

And as others have mentioned temperature is a big deal. Hanging onto a part for 20 seconds will move the part.

 

[Conrad Hoffman]

I've always assumed that if length isn't specified, the number applies over the whole length or surface.

Something I've found to be terribly fussy is the ends of v-blocks. I want to be able to clamp a shaft or pin and grind the end of it square. Many/most v-blocks have more error than I'd like for that. It's probably better to use a pair of decent 1-2-3 blocks.

 

[JST]

..............................

And as others have mentioned temperature is a big deal. Hanging onto a part for 20 seconds will move the part.

Same for the camelback straightedge. Even 10 sec of your bare hand will make it a banana for close tolerances. Make some wood handles for it.

 

[marka12161]

I found scraping a pair of v blocks a great project. Had to ask for advice here on how to get the v centered and at 45deg. In my non Moore quality room, I’m not hitting half a tenth. Nor can I realiably measure to that fussiness with my grade A plate and my gack. What condition is your plate? How stiff is your method of holding your tenths indicator and is it 3 pointing on plate or is there rock? Many other questions.

2 tenths over 12” is something I can reliably hit (I’m an amateur). I have done 60 ppi on two smaller projects and never want to do that again. I’m sure with practice you can do the same or better. Just please listen to people smarter than me and don’t chase better than needed accuracy.

Thanks for the feedback. I had my granite plate inspected and certified to grade A about three years ago. Good enough for this hobbiest. I'm holding my indicators with surface gages (ground bottoms but stoned to remove burrs before use). I do have a nice three point indicator holder but I need to come up with a way of attaching the indicator. I'm learning a lot fast. My scraper is shop made and, as you would ecpect, i've got some changes in mind for my next attempt. So far for me, the most challenging thing has been managing the thickness of the cannode so it doesn't smear yet spots adequately. I think I've got that figured out. I'm having great fun. I find it to be a bit like eating potatoe chips (just one more...).

 

[michiganbuck]

One and two-tenths in a distance of 3 inches are usually ok for general shop job work but are not close for gauge work or precision fitting. A very good grinder might be capable of a few tenths in a few feet, but with wheel-breakdown magnet suck down and heat. that is tolerance is not always easy.
*Scraping is often on a relaxed part so transferring the accuracy of the almost perfice checking fixture to the part is closer than grinding
One serious often ignored problem with high-precision scraping is that thoughtless assembly can lose the intended quality of the scraped part. One example is a surfacer grinder long travel top table. I have a torn-down SG in my shop now that has a visible failure of the scraped iron with oil ways. One can easily see that the original scraping is still like the original at the out ends...and worn in the middle. The fault cause was an overly tightened magnetic chuck.
The scalps in the middle are not completely gone so the belly wear may be .0002 to .0005 (about).

Rough idea of how much pressure might be on 2 screws overtightened mag chuck. The clamp load is roughly 75% of the proof load.

Proof Load Tensil Strength For Grade 2, 5, & 8

Tensile Strenght: Proof Load Tensile Strength for Grade 2, 5, & 8,

www.almabolt.com