How to Produce A (Deep/Narrow) Precision Slot

[carrdo]

Hi All,

Guys (and ladies)(I guess there are a few).

I don't know why I am telling you all of my secrets. Nobody taught me this; it was only learned through the school of hard knocks. It is probably one of the most difficult of all machining operations. There are many ways to do it wrong and only a few ways to do it right.

One project which I had was to produce a precision narrow deep slot. By precision I mean a slot that was truly rectangular, was not bell mouthed, was straight and on size to 0.001" or less with a smooth surface finish. The slot in question was 1/8" wide and 1-1/8" deep. It was for a replacement scriber carrier on a Starrett 6" vernier height gauge which I have used for decades. This height gauge is a beautifully made precision tool so do everything to keep it that way. Don't wave it around or drop it, bump or nick it by throwing it in with any other tools or metal pieces, keep it scrupulously clean and lightly oiled and it will last longer than you will. But I digress. It is not cheap and condition is everything but is it truly expensive if it will last more than a lifetime and as well keep all of its built in precision...?

To start, material selection is very critical. No ordinary steel is right for this job. Starrett makes a special toolmakers steel called gauge stock which is a special alloy steel which has been heat treated (and then ground flat in various rectangular sizes) to virtually eliminate any internal stress so that the material does not move when machined even though on a few occasions I found that even this material bell mouthed a bit (about 0.002") when making these slots. To be fair though, it may not have been a true Starrett product as I have purchased other "gauge stock" products on occasion. 1144 Stressproof steel is a possible option but it only comes in round bars and I have not tried it.

Why any normal steel is unsuitable is because of the internal stress inherent in the steel due to surface cold rolling or cooling where the exterior of the steel cools first before the interior does. Theoretically, you need an infinite cooling down period to avoid the aformentioned condition but since this is not possible, other methods have to be employed. Cutting, shaping or bending simply redistributes these internal stresses but does not eliminate them.

to be continued.

 

[Proxule]

Great, I am all ears. I have a similar height gauge that is missing the scriber portion, So perhaps I can learn some thing
Thanks for posting!

 

[Tom O]

Great, I am all ears. I have a similar height gauge that is missing the scriber portion, So perhaps I can learn some thing
Thanks for posting!

Is that the one off marketplace?

 

[Proxule]

I bought this several years back. Sitting idle until I have a genuine use for it.

Are you referring to a thread or comment I missed ?

Thanks

Is that the one off marketplace?

 

[Tom O]

Nope there was one on marketplace missing the same parts so I was just wondering.

 

[Susquatch]

Starrett makes a special toolmakers steel called gauge stock which is a special alloy steel which has been heat treated (and then ground flat in various rectangular sizes) to virtually eliminate any internal stress so that the material does not move when machined even though on a few occasions I found that even this material bell mouthed a bit (about 0.002") when making these slots. To be fair though, it may not have been a true Starrett product as I have purchased other "gauge stock" products on occasion.

Stress relieving before and after machining is a fairly common practice for critical parts. When you think about the forces and temperatures associated with machining, it's not surprising to me at all that you experienced 2 thou of movement. The good thing is that you knew it happened and were able to deal with it.

 

[carrdo]

Hi All,

To continue.

First rough out the blank and mill it dead rectangular, square and parallel leaving it 0.015" oversize. This is important for if you get this wrong all subsequent machining operations will be wrong also.

Ink and layout the central slot leaving the layout lines 0.010" smaller than the finished slot width on each side with the bottom layout line being the full slot depth.

The secret is being able to first bandsaw out as much material as possible up to (but not over) the layout lines. Use a fine (32 wavy tooth) blade, the sharper the better. Again, the secret is not to push the bandsaw blade, just let it cut by itself and use one's fingers only to steady and guide the blank against the fence of the bandsaw. It takes forever in this material but if the procedure is followed, the blade will make a straight cut. Good lighting is essential.

Cut down on each side of the slot and then the bandsaw blade can be used to remove the centre strip piece (if any is left). One can also use the bandsaw blade somewhat like a milling cutter to roughly square the slot end if used carefully.

What one is trying to achieve here is to pre remove as much metal as possible before any of the finishing operations are attempted.

to be continued.

 

[carrdo]

Hi All,

Moving on to the finishing operations. One needs to have a new or freshly sharpened side tooth slotting saw not just one having phepheral cutting teeth as an excellent finish is important here. I used a 7/32" width side tooth slotting saw followed by a 1/8" thick side tooth slotting saw as seen in the photo just to make sure each saw had the minimum amount of metal to remove which again minimizes, to the most practical extent, the cutting forces on the part and the saw teeth.

Note that one needs a large diameter cutter for such a deep slot. If the holding arbor has a 1-1/2 inch diameter backing flange and the overall saw diameter is 4", this leaves only (4" -1.5")/2 = 1.25" free cutting depth which is just about the minimum to have a bit of slot depth clearance. Also, I don't use questionable offshore cutters for operations like this.

But wait, there is more! Don't reach for you cutting tools yet.

To start with the basics - your mill should be checked for tram and the mill's table, holding vise, vise jaws, etc. should be free of any nicks or dings and everything aligned square. I am also presuming that you know how to set the saw central to the work. Again, I use my trusty Starrett 6" vernier height gauge for this height setting operation.

Second, the holding arbor is equally as important as the cutter. The arbor must hold the saw at a true 90 degrees the the vertical centreline of the mill's spindle without any wobble, tilt or looseness. For example, if the backing flange face on the holding arbor has been ground to say 0.0005"(1/2 thou) which is a pretty tight tolerance square to the spindle centreline, there will be significant wobble at the OD of the 4" diameter saw as (0.0005" x 2" [saw radius] = 0.001" wobble). But it gets worse as the 0.001" wobble has a high side and a low side. The saw will now cut 0.001" too much on the high side of the wobble and 0.001" too much on the low side of the wobble giving a total oversize cut of 0.002".

So what we are looking at is a top of the line arbor whose flange face has been ground to at least 0.0002" square to the vertical centreline or better for a job like this. They do exist but expect to pay $$$. I have checked the arbors which I have using an 0.0001" indicator and with the arbor shown it gave a wobble runout of 0.0003" at the cutting edge so 0.0006" total wobble which was OK.

The rest is just good machining practice. Use plenty of cutting oil, keep the speed of the saw low (at the saw OD) and use the table stop on your mill to ensure a soft landing at the end of the slot. Again, total control of the tip cutting forces and absolute rigidity are a must.

PS: Upon reviewing my notes, on one of the height gauge scriber carrier slots which I made, the material, after slotting, bell mouthed a total of 0.006" (not 0.002" as stated previously) which is completely unacceptable if, indeed, it was Starrett gauge stock but as I indicated earlier....

In this job one "closes the rectangle" with the scriber holder headpiece which is riveted in place (as per how Starrett did it but that is another story). All of the parts are surface ground to finished final dimensions as is the headpiece and some of the rectangle spacers shown in some of the photos in this thread posted previously.