How to correct mistakes

[Mr Ron]

As an amateur machinist, I try to machine a part to within a thousandth of an inch of it's design dimension. I can take too much off or not take enough off. How to correct this? If not enough has been removed, it becomes a matter of going back over the part and removing more material. By doing this, I take a risk of ending up taking off too much. At this juncture, I can scrap the part and start all over. My projects are all one off, so interchangeability is not a problem. Rather than starting over from scratch, I will make a design change using the arrived at dimension of the part. Although having deviated from the original design dimensions, the parts will still work as intended. I'm not happy when that happens, but I am only an amateur. I try hard to maintain a dimension, but I don't always do so. To a perfectionist like Harold V, it would not be acceptable.

What do you, forum members, both experienced and amateur think about my approach to machining? I'm sure there are amateurs out there who can produce extremely precise work, but I am not one of them. At 82, I am running out of time and realize I will never get to go to that machine shop in the sky.

[whateg0]

In my shop, everything is a "prototype"! This is what I was alluding to in my "close enough" thread. I think most of us on this forum prefer to make perfect parts. When there is no drawing, what works just works. I often joke about the products we make at work. If one day somebody discovers that something doesn't meet the published spec, our engineers have been known on more than a few occasions to widen the tolerance for that spec, or often just remove the spec and replace it with a "typical" value.

[Benta]

This is why you have calipers and micrometers.
I always go towards a dimension, and before it's reached fully, I stop and do a check measurement to the best precision possible.
You cannot rely 100% on the precision of the lead screw, and workpieces and tools always flex a bit while working.

Based on the measurement taken just before reaching the end dimension, you can compensate, and do the last cut to very high precision.

[toddalin]

Depending on the part and application, I've used paint as a shim. Best to cook it on at low temp first.

There is also "flame spraying," but that's $$$.

[Harold_V]

Based on the measurement taken just before reaching the end dimension, you can compensate, and do the last cut to very high precision.

That's not reliable unless you learn to take multiple cuts of equal depth. When finishing with a tight tolerance, I like to leave .03" of material (assuming the material machines without tearing). I pick up, take 1/3 of the remaining material, measure, take ½ of the remaining material, measure, then take the part to size. That way you maintain identical tool pressure and avoid surprises.

H

[tornitore45]

What Harold said. Also avoid removing the part if you can accurately measure in place. Re-clamping or re-chucking the part can invalidate your dial references. Same about dial movement, keep going in the same direction. Backing up and returning to the same spot is no guarantee.

[GlennW]

If you have a bit of flexibility as far as size and have a bit of difficulty machining to an exact dimension, what some do is make the part that is the most difficult to size first, which is usually a bore. Then when you make the part that fits in the bore you can get it as close as you can and then finish it up with a bit of work with emery paper or similar until it fits the way you would like it to. You may also end up with a better finish as a bonus.

It may take a bit more time, but so does making the parts over again!

Be sure to drape a cloth over your ways though to catch any stray grit.

[Rolland]

Based on the measurement taken just before reaching the end dimension, you can compensate, and do the last cut to very high precision.

That's not reliable unless you learn to take multiple cuts of equal depth. When finishing with a tight tolerance, I like to leave .03" of material (assuming the material machines without tearing). I pick up, take 1/3 of the remaining material, measure, take ½ of the remaining material, measure, then take the part to size. That way you maintain identical tool pressure and avoid surprises.

H

That is exactly how my father taught me to machine a piece, he called it "sneaking up".

[Frank Ford]

Don't be too hard on yourself! Thousandths are pretty small things, and most folks don't know if you've lost a few here and there.

Ok, seriously, I see nothing at all wrong with making things looks good and work well, even if they doesn't meet your original specifications or ideal. I can't tell you how many projects I do where that simply doesn't matter. Sometimes I do need to hit dimension with real accuracy, but when it's not important, I simply don't have the time and emotional horsepower to get all worked up about it.

If people like my stuff and I like my stuff, that's good enuf for me. . .

[Harold_V]

Yep! What Frank said. The only real exception is when you work to specific dimensions, and the work will be accepted or rejected by a third party. They won't have the luxury of accepting the part, or assembly, because it functions---as the print rules.

When one works in the trade, the luxury of doing as you wish isn't always an option. It's safe to say there's two sets of rules. Which set you choose to uphold is a personal choice. Mine is to work to the print, as that's how I was trained. It's not unusual for me to toss parts I screw up (even when working on my own projects) and do them over. That's how my head works.

H

[Mr Ron]

Sounds like a lot of good advice. I always wanted to be a machinist all my life, but my parents had other plans for me, like college. I then got caught up with the reality of life and being a machinist was never to be. Now as an amateur, I can realize some of that dream.

[whateg0]

Harold,

I think that "two sets of rules" thing is the important part of these discussions. Sure, when working to a drawing, it's important to "get the numbers right". Much of what we do as hobbiests is dictated by other things, such as material on hand. Just a couple nights ago, I was making a new clamp to hold an incorrect scriber on my Starrett height gauge. I didn't have a drawing to work from other than a quick sketch on a piece of scratch paper so I could do some math. Just measured the parts that had to fit together and made it all fit. Once I had a rough idea of what I wanted to do, I looked in the drops bin by the bandsaw and found a piece of steel that looked like it would work, then made the part fit the stock. Turned out great, IMHO.