First attempt at precision work : success, but...

[Cowbell]

... I think I generated more questions that the ones I figured out

Preface : Beginner, working with a humble Grizzly mini-mill. After tinkering around for a month or two, I built a little homebrew flood cooling system and prepared for some precision work.

My goal at hand was to machine a block of aluminum to 0.751" wide, on a bar 0.5" thick, ~4" long. At the end of the day, I got it to < +/-0.001, but man did I come up with some issues.

Issue 1 : How flat is my milling table? I trammeled (sp?) my table with the vertical head. Using a dial indicator with 0.0005" gradations on the spindle, I was able to seemingly get both ends of the table to <= 0.001". I spent a ton of time rotating that thing back and forth and I can't imagine getting much closer with the tools I have to work with. Now the weird part - if I run the dial indicator on the right half of the table, it stays within that 0.001". However, once I cross over to the left half of the table it starts to diverge quickly. Several thousandths over the 5 or so inches I can measure. What gives here? First off - is this level of precision normal for a low end mill? Second - what might be the cause of one half of the table to run out so much while the other half stays pretty even?

Issue 2 : How flat is my vice? I've got the low end crappy Grizzly vice. When I snug that guy down on the "good" end of the table, the run out is severe. Multiple thousandths over 2+" of horizontal travel. The only solution I could come up with was to fold over a piece of paper under the low side to raise it up a bit and then adjust the tightness on the T-nuts until it ran pretty close. Clearly, this is not a great solution. A vice that lousy isn't particularly usable. Is there a fix for this? I tossed around the idea of milling the rails of the vice myself, but I'm not entirely sure it's ready for potential destruction by me I'm not entirely opposed to buying a higher end vice, but I'd like to be sure there's actually a problem with the vice and not the user. Any advice?

Issue 3 : Parallelism. Even though I got my bar darn close to the desired dimension, one side is ever so slightly non-parallel to the other. If I stick it in the calipers, one side is flush, the other is just a hair off. In theory, if I endmill one side, then if I flip it over on to a perfectly flat working surface, endmilling the other side should result in perfectly parallel surfaces, no? I'm assuming this isn't working for me because I don't have a great technique for mounting the piece perfectly flat. I'm laying down two parallel bars on the base of the vice and balancing the workpiece on them as I tighten the jaws. But it never feels very snug. When the jaws tighten against the workpiece, the parallel bars get tilted ever so slightly. I can adjust and readjust, but it never quite comes out right. The end results confirm this. Anyone have any suggestions on getting things nice and flush?

Issue 4 : Backlash. My mill has some serious backlash that results when I tighten down the table motion screws. This is particularly evident on the vertical feedscrew. If I fine adjust right where I want, as soon as I tighten the motion screw, I'm looking at up to several hundredth's of an inch of offset. This is present on the horizontal feedscrews, but not quite as pronounced. The only thing I could come up with to get by this was to pre-tighten the motion screws right before I reached my range and then crank the feedscrew from there. This works fairly well, but it feels very much like a bad thing to do. If the motion is locked down, I'm putting a ton of stress on things as I force the feedscrew around. What the heck is the solution here? The feedscrew is accurate enough, but locking it down when done just blows any precision waaaaay off. I had to make 4 passes over the piece to get it down from 0.865" to 0.751", since each time wasn't quite where I thought I'd end up.

Welp, there's my questions. I would really appreciate expert advice here. I had a hoot cranking this thing out today (particularly since I got to see the flood cooling in action - VERY neat), but it felt very "hacky" instead of like truly precise work.

[JPMach]

First off I would suggest a better mill, but short of that the first thing I would do is check the level of the machine with a precision level. Check side to side level in the far front and far back of the table travels and check front to back level on the right and left side of machine. This ensures the machine is not twisted, a table can read level in both directions in the middle yet be twisted from uneven leveling scres adjustment.
Once you have done this then retram the head. If you can not get the head trammed within a thou then there is something else wrong.
As far as the table being flat, I doubt it. I bought several chinese vises when I started my shop to save some money at first, I had to spend a couple of days surface grinding them done so that the would sit flat without rocking and so the top was flat to the bottom ( I kick myself for not just buying Kurt ).
As far as backlash, a good machinist can work around that. One of the things I had to do in school was machine an inside square pocket to +-.0003" on an old bridgeport with around .030" backlash.

JP

[jkeyser14]

Do you have access to a shop w/ a surface grinder? If so, you could grind your table, bottom of the vice, and vice jaws to make sure they are pretty flat (I'm assuming they aren't aluminum).

[JPMach]

When you get the mill head trammed, try spinnin an indicator over the ways on your vise and see if the vise is in tram. If not then find someone with a surface grinder.

JP

[wizard]

There has been mention of tramming the head in but that is jumping the gun. Tramming is only usefull if the all the components below the head are right. You have left me with the impression that this might not be the case.

What you need to do is to take a good dial indicator on a solid mount and check for proper operation of the X & Y first. With the gage stand solidly mounted you should be able to move the X & Y a considerable distance and see little in the way of deviation. I'm not going into detail here as I don't know what the spec are for the mill, I'd assume they are loose simply due to the price paid.

Also I don't know what type of dial indiicator you are using. But some are sensitive to the driection of travel relative to the needles axis of rotation. Also you need to verify after each measurement that you gage did not move relative to the starting position.

Once you are sure the X & Y are good then you can think about tramming the mill head in. On larger mills at a minimal you should be able to get less that 0.001" deviation with a 6 inch radius. Now I'm not going to say that that is possible on this type of mill but you ought to be able to get close.

Thanks
Dave

[badhoopty]

To go back to an earlier post. I would be sure that the machine is good and level first, then follow Wizards instructions for finishing it up. I have worked in machining precision parts for over 16 years now, and I started my career on a bridgeport handmill. My second suggestion here would be to not buy cheap equiptment. You would be better off buying "used" brand name stuff than buying anything with the a name even remotely resembling "GRIZZLY." The only reason I knock Grizzly pretty hard is that we have a small grizzly hand lathe that was meant to do some secondary hand work from our cnc lathes and it isn't even good enough to do that. It "is" pure junk.

BH