Is the block really down on the Parallels?

By R.G. Sparber

January 16, 2007

Copyleft protects this article. You are free to take it but not to claim you wrote it. Fair enough?

Have you ever started with a nice piece of square bar stock and wanted to reduce its thickness? You put down your favorite mill vise, true it up. Then you drop in a parallel. On top you place your bar stock and tighten the vise. After a few passes with the end mill, you remove the bar to discover you now have a wedge. Not a happy moment. So you put the bar back in the vise, find the lowest point, and machine across it again. Maybe this time it really is flat or maybe it is just sloped in the opposite direction. Not sure what is going on but leaving the shop for some mindless TV show is starting to sound appealing.

With just a few pointers, you can avoid these unhappy moments. It all boils down to some simple math: 1 + 1 = 2.

Start with the vise. You should have used your Dial Test Indicator (DTI) to set the fixed jaw true to the X axis but did you also verify that the ways of the vise are parallel to the table's surface?

I touched down with my finger DTI on the ways in the foreground and slowly moved to the ways in the background. Any deviation from zero that you see on the DTI will be an error in your cut surface. If you do see an error, remove the vise and do a good cleaning of all surfaces. Using a strong light, inspect for bumps in the table's surface and in the underside of the vise. A bump 1 thou high can show up as a 1 thou error across the vise ways. Then run the DTI along the mill table top to be sure it is flat. If the mill tabletop is not flat, you may have to scrape or lap it. That is beyond the scope of this article. Last year I spent a not-so-fun day with a lapping plate making my mill table flat. It was worth the effort.

Assuming the vise ways show zero all the way across, you are ready to place your parallel. I try to use only one parallel because it is very rare that you can firmly rest on two at the same time.

Since your vise ways are true, you should see zero all the way along the parallel. If you do not see zero, remove the parallel and clean all surfaces. I use WD-40 from a spray bottle and then wipe down all surfaces with toilet paper. A strong light can often help you spot tiny particles of debris that will lift up the parallel. The bottom line is that given that the vise ways read zero, and the parallels are in fact parallel, you must read zero along the top of the parallel. Inspect and clean until this is true. Do not overlook the possibility that the parallel is not true. If in doubt, mic its height at various points along its length. Also place it on a second parallel to verify no warpage. There should be no light showing between two parallels placed together.

When you read zero along your parallel, it is time to bed down the block to be machined.

Before you put the block into the vise, measure its thickness at each end. First I will discuss the case of the block being of uniform thickness. You inspect the surface that will be in contact with the top of the parallel. It must be clean and free of bumps. If in doubt, put down a piece of emery paper on a flat surface and move the block over it until dead smooth. Then place the block down on the parallel. I use a piece of newspaper between the movable jaw of the vise and the block. This greatly improves the grip.

Gently tap the part down with a soft face mallet while you tighten the vise jaws. Do not go crazy with the mallet because you can bend the block a thou or two. Believe me because I did it last week. With the block bedded on the parallel, again run the DTI along its surface. If it reads all zero, you positively know that the block is in full contact with the parallel and the parallel is in full contact with the vise ways. If you do not read zero, remove the block and figure out why. For me, it is often one tiny fleck of swarf. The fact is that there is a reason why you are not reading zero. Do not proceed until you eliminate the source of error. Once you read zero along the top surface of the block, proceed to make your one or more rough-cuts followed by your finish cut. I believe you will be happy with the results.

What do you do if the block is not of uniform thickness. Not to worry, the math still applies. Say the left end is X thou higher than the right end. As you run your DTI, expect to see this same difference as you move from end to end. If X is a thou or two, you can get away with using a finger type of DTI as shown here. Try your best to have the force on the DTI's finger be perpendicular to the finger's length. This will minimize DTI error. If the difference is more than a few thou, use a plunger type of DTI, which is better suited to measuring displacement. Alternately, if you have a Z axis digital readout, you can use a finger DTI set to zero on one end of the block. Zero your readout. Then move to the other end of the block and move the spindle until the DTI again reads zero. The readout will show the difference. When you see the left end X thou higher than the right end, the block is solidly bedded on the parallel and vise ways.

Thanks to Brian Lamb on the proper use of a DTI.

Your questions and comments are welcome.


Rick Sparber