This thread is just about to reach "epic".
If there was any doubt this post ought to push it over the edge.
Still, I have asked it a couple of times. No one has answered. How are many of you actually checking your work after all criticle areas are finished?
I check what I can with the instrumentation I have. I'm a hobbiest doing it for myself and family. I don't have any long reach DTIs, but I don't think much of them in any event (they aren't rigid or mounted rigid +/-0.0001" so how the heck can they measure that accurately bouncing over rifling? Answer: They can't.) so I don't regard that as a handicap:
- I put the Grizzly rod into the barrel at and and inch or so past the throat and check to see if it is still where I had it when I aligned the blank in the headstock. I understand the tolerances in the Grizzly rod approach. The DTI is supported off the tool post which is a heck of a lot more rigid than a magnetic base, the rod is almost bore size, and it's held against one side by a thread and 2oz weight - it repeates beautifully. I use the mag base to guide the thread - it works good for that.
- I use the DTI supported off the tool post on the back end of the chamber to see if it is running true to the lathe axis.
- I check the chamber depth one last time with the Go-Gage and a depth mic.
- I check the chamber diameter at the breech end to see how it compares to the reamer.
- I look at the throat area with the borescope to see if it is centered.
The barrel blanks don't seem to move, but that isn't surprising considering the light cuts I'm taking.
I've not had any oversize chambers using my floating pusher, and so far at least, the rear of all the chambers has been within +/- 0.0002" and the throats are centered within the tolerances permitted by the bushing fit to the bore. For me, that's good enough.
I have spent quite a bit of time in metrology labs working the the experts to set up measurements in my former life. It is a whole lot harder than most folks would think to make a measurement that is repeatable to within +/- 0.0001". Not gonna happen with a mag base supported long reach DTI bouncing over rifling. The rigidity just isn't there. Period.
I can see a significant difference between holding my DTI off a holder that looks like a cutting tool blank clamped in the tool post and on the mag base. In theory and practice, if the force to deflect the tip and get the measurement isn't much much much less than the force to deflect the setup by the accuracy delta desired, the measurement is not going to have the desired accuracy.
When I first got my brandnew DTI and tried it out I thought it didn't work. Then I supported it off the tool post and came close to throwing the new magnetic base in the trash - it is useless for measurements that need to be less than +/- 0.001". DTI was as expected, the mag base was not rigid enough.
After reading how many are setting barrels up, I can see "stacked variables" permiating many.
And some approaches are fundamentally flawed, some more than others, in addition to tolerance stacking. If the setup isn't theoritically correct for the general case of the part to be machined, results are determined, at least in part, by chance properties of the particular piece. The goal is to get the desired result every time, not just when the stars are aligned.
In the Machine Shop world, you have to check and verify. You simply cannot assume that a given set-up is performing as it should.
About two months ago, an old friend, (retired ploice officer), who has got into gunsmithing, brought an action by that he has "trued". He had invested in that system that uses an arbor and a tap to straighten up the threads. I forget which brand.
He asked me to see how well it did. I machined a dummy barrel tenon, screwed the action up on it, and placed my checking arbor in the end. The arbor ran out .020.
I explained to him that his threads and action face still were not truly straight with the bolt way. I tried to explain to him all of the variables that can stack up when using a rig such as what he used. The major one being that the entire concept was flawed from the beginning.
Roger that! I read a description of the tap systems and figured there must still be suckers born every minute if they sell. There is no way the tap system, which is fundamentally flawed from the start, is going to be anyplace close to as accurate as single point chasing the threads in a lathe after doing the proper alignment. That said, it ain't as easy as it might look to get it right. I've only done it a few times and it took me hours to get ready for a few minutes of making chips.
Unless you are actually checking your set-ups and procedures, you are simply painting by the numbers..........jackie
To which I'd add measuring the completed part as your friend asked you to do. Good on him for following up to check his work! I wish it had come out better than it did, but considering the tools he used, it could have been even worse. Truth be told, with out data on how it measured before he worked on it, there is no way to know if his efforts resulted in an improvement or made it worse.
Your machined stub method for checking receivers ought to be pretty darn good if the stub is used right where it is machined and not removed from the machine before use. The machine is rigid, the instrument can be supported on a tool post resulting in a setup that is close enough to perfectly rigid relative to the forces involved in taking mesurements that the difference is undetectable in a practical sense.
Fitch
