Hole Layout experiment

[raferguson]

Over the years, I have laid out holes by hand, using scribe marks, a center punch, and drills. I have been less than satisfied with the results, as the holes are often visibly off. As I recently acquired a mill-drill, I decided to see what kind of results I could get, using different processes. I am definitely not a skilled machinist.

I used 1" by 6" by 1/8" hot rolled steel stock for this experiment, and drilled using my drill press unless otherwise noted. The stock was held in a vise, which was not clamped down. Using a digital caliper, I measured the metal between the holes, and between the holes and the edges of the stock. When I center punched, I generally did a first punch using a small automatic punch, and then made the punch bigger with a larger automatic punch.

I started with what I have mostly done in the past, which is using a machinist square to mark the metal, center punch, use a 1/8 inch pilot drill in my drill press, and then a 3/8" twist drill. The standard deviation of the hole placement error was about 0.016 inches, meaning that 95% of the holes would be within 0.032" of the desired location, not so great.

I did the same thing again, using a center drill instead of a 1/8" pilot drill. The results were slightly better, with a standard deviation of about 0.012".

Another time, I used layout fluid rather than scratching marks in mill scale, using a center drill. I also sanded off the mill scale, to make the marks easier to see. The results were better, although I made an error and one hole was 0.030" off, hence I am not reporting a standard deviation. (I really need to repeat this test).

Then I took it to my mill-drill, using a 3/8" mill cutter. The results were dramatically better, with a standard deviation of 0.003".

What have I concluded from these experiments?
1. It is more accurate to start a hole using a center drill rather than a 1/8" drill bit.
2. Using layout fluid rather than marking on mill scale produces more accurate hole placement.
3. The mill-drill produces more accurate hole placement. (Of course).

Do these results make sense? The results seem as I expected.

Richard

 

[dulltool17]

They are completely logical conclusions, supported by decades, if not centuries, of experience.

 

[benmychree]

Good hole location starts with an accurate layout; your layout is good enough for work not requiring good accuracy. For accurate layout, I use a vernier height gage, and use a prick punch to locate the holes; a very sharp punch dragged along the layout line can feel the intersection of the line, then straighten up the punch and lightly hit it with the hammer, observe if it is right on, then hit again a bit harder, then with a center punch to make a larger impression, I then would use a center drill to pick up the punch mark. If the prick punch mark is a bit off center, one can tilt the punch over and push the impression whichever way it needs to go, then punch vertically again.
Hole locating by machine movements is likely to always give better results than hand work, but I like to do an accurate layout to avoid mistakes in counting turns on the machine, not having any DRO for my mills.

 

[raferguson]

Help me understand the prick punch discussion. I have what I call center punches, the smallest of which is a Starrett automatic punch. The end is about 0.075" in diameter, and appears to have a 45 degree angle (so 90 degree total angle, hardly a sharp prick punch). Do I need punches that are much sharper than that? I could make whatever punch I want, I suppose, or just buy one, but not clear that whatever I buy would be very different from what I already have....

How is the vernier height gauge used? I get the idea of making scribe lines that I can feel, though how to scribe that deep, accurately, is unclear to me. I am particularly unclear on how I could find the intersection of two scribe lines.

I have been using the scribe that came with my machinist square, have not tried to sharpen it. I have sometimes used a carbide scribe, which I suppose could make a deeper mark if I press harder, but then I worry about the ruler slipping under the pressure.

dulltool - I had to laugh about "centuries of experience".

 

[benmychree]

The vernier height gage is used on a surface plate, you set your job up on whatever holding means necessary with the surface to be scribed vertically, set the height on the gage and scribe with a sharp scriber attached to the gage. Cross lines are scribed in another set up, and the junction is prick punched; the prick punch has an included angle point of 60 degrees, making it easier to accurately locate a center point. If it is very sharp it is easy to locate quite accurately. Sometimes, the punch will stand in its own mark after being hit with the hammer. Personally, I do not like or use automatic center punches, I can do more accurate work with the prick punch.

If you drag a prick punch along a scribe line at an angle, when you come to a cross line the punch will drop into the scribe line, you then straighten the punch up and hit it lightly with the hammer.

Note that when doing a layout with a square and scale, the scriber always makes a line the is a little bit away from the rule, not conducive to accuracy; using the height gage removes that inaccuracy.

 

[RJSakowski]

I make my prick punches from old chain saw files. The angle on the tip is about 45º and they are sharpened so they will bite into m finger nail if I lightly push them. The struck end is heated to a dull red to draw the temper. I also make scribes from the files. I make the angle about 30º and sharpen like the prick punch. Since they are not struck, you don't have you don't have to drawe the temper.

 

[matthewsx]

The surface plate method works great, gave me an excuse to buy a height gauge too.

A friend did something different last night though, he layer out the holes he needed in CAD and 3D printed a template which he then applied to the steel he needed holes for. Apparently it worked out quite well, the project was for linear rails on his CNC router.

 

[mikey]

Lots of ways to do this and it really depends on how accurate the hole placement must be.

You can scribe lines and use a prick punch followed by a center punch and that will usually get you within a few thou when you use that mark to center the work with a center drill. I mean that the work piece is allowed to move, which allows the center drill to move the work so that the drill itself pulls the work on center of the punch mark. Then you clamp the work to the drill table and drill the hole. If your work piece cannot move because it is locked in a vise then this method is not that accurate.

You can also scribe intersecting lines and use the sharp point of a wiggler set to visually get you on center. This usually gets you within a thou or two. You can do the same with a center drill by just touching the intersection with a center drill and then move the work or the work table around until you are visually centered on the intersection.

If you need to get really accurate hole placement, lock the work in your vise and use an edge finder and the X-Y handwheels of your mill to get the drill precisely located using your desired coordinates. This can get you within a thou or less of dead center. Then either use the tip of a center drill or a spotting drill to create a starting divot, then drill with a drill. Use good drills held in an accurate chuck with proper technique and you'll be pretty accurate.

How you drill also affects your accuracy. If you use the very tip of a center drill or a spotting drill to create a starter hole, this allows the drill that follows those drills to start at the dead center of the hole those drills make. Then you need to feed the drill slowly until the tips of the flutes are below the surface before you apply more feed pressure; this allows this initial hole to guide the drill accurately. Drill this way and you'll be on the mark with much greater success.

 

[C-Bag]

Over the years, I have laid out holes by hand, using scribe marks, a center punch, and drills. I have been less than satisfied with the results, as the holes are often visibly off. As I recently acquired a mill-drill, I decided to see what kind of results I could get, using different processes. I am definitely not a skilled machinist.

I used 1" by 6" by 1/8" hot rolled steel stock for this experiment, and drilled using my drill press unless otherwise noted. The stock was held in a vise, which was not clamped down. Using a digital caliper, I measured the metal between the holes, and between the holes and the edges of the stock. When I center punched, I generally did a first punch using a small automatic punch, and then made the punch bigger with a larger automatic punch.

I started with what I have mostly done in the past, which is using a machinist square to mark the metal, center punch, use a 1/8 inch pilot drill in my drill press, and then a 3/8" twist drill. The standard deviation of the hole placement error was about 0.016 inches, meaning that 95% of the holes would be within 0.032" of the desired location, not so great.

I did the same thing again, using a center drill instead of a 1/8" pilot drill. The results were slightly better, with a standard deviation of about 0.012".

Another time, I used layout fluid rather than scratching marks in mill scale, using a center drill. I also sanded off the mill scale, to make the marks easier to see. The results were better, although I made an error and one hole was 0.030" off, hence I am not reporting a standard deviation. (I really need to repeat this test).

Then I took it to my mill-drill, using a 3/8" mill cutter. The results were dramatically better, with a standard deviation of 0.003".

What have I concluded from these experiments?
1. It is more accurate to start a hole using a center drill rather than a 1/8" drill bit.
2. Using layout fluid rather than marking on mill scale produces more accurate hole placement.
3. The mill-drill produces more accurate hole placement. (Of course).

Do these results make sense? The results seem as I expected.

Richard

In a way they are for just practice arbitrary holes but the whole process got way trickier when I had to try to match an existing pattern. Measuring super accurately became the problem when I had mastered the bluing and layout that benmychree described. Not until this came up here did the Sorensen Center-Mike get mentioned and that was the next level for me. Then it came back to my limits on the mill/drill and that got dialed in with the addition of a DRO. For me it was continuous waves like you describe where the whole process would get better from beginning to end with the result being better each time and what you describe was the second or third wave as I was forced to get better and more accurate in measuring, layout and execution on the mill. It sounds like you got better results than I at the stage you are.