To accurately make that part...teh method described above by Derek is the way to make this part and absolutely no micro rotation of the part is needed or wanted.....
Can run this either horizontal or vertical....Grip the body on the OD. using the chuck if possible or a collet on the shoulder just ahead of the hex for turning...
If held via the chuck no tail stock is needed
Anyhow align the part so that it is running true with the "X" axis, and has minimal runout on the OD.
Set the part so that the spindle is on center with the center of the part rotation using the "Y" or "Z" axis depending on if you go vertical or horizontal...
Select a good end mill, you want something relatively small in diameter so that it will fit between teh opposite tooth of the part as you are cutting and to not need to go too deep past the tooth depth in order to
make the side flat.
If the tooth is 6mm wide you simply need to move the spindle from the centered position by 3mm and 1/2 the diameter of your tool....then cut one side of all the teeth....
indexing each, then cutting ....Personally i would work up to the final dimension ...originally moving further than the finished calculated position...Finish all one side of all 10 teeth.
Then move the spindle to the opposite side and work to the same offset only on the opposite side of the part centerline.....
As you approach the finished width of the teeth (as measured) finish to size simply by continuing to cut by moving the spindle closer to the center till the final size is reached....
If you carefully cut the first side, you can finish to size on the second side...a small difference between the exact cutting position on each side will not affect the part.....
Also...this is a tool and you should not strive to make it fit too tight in the drive notches of the nut....give it some clearance, won't hurt the function and will make your life easier....
For 10 divisions on your dividing head...the indexing is 4 turns in any hole circle...Formula is : N= number of divisions. T= Turns of the crank
Given a 40:1 ratio in the head (pretty standard)
Then...40/N = T Or 40/10 = 4 that is 4 full turns on any hole circle.
A final note. its good to be aware of accuracy and strive for quality work, but gilding the lilly is costly and often the results are less than expected....this is not a part that demands micron accuracy...Remember to not make it fit too tight on the teeth...if it does not fit easily once you take the part off the machine,giving it more clearance later is much harder.....
Cheers Ross
Hi Ross,
thanks for the reply.
Understood, the method you and Derek explained is clear to me.
Thing is, like I've mentioned before, that the whole discussion was not on making the specific part, which, indeed has no strict specs in terms of accuracy.
I wanted to know, theoretically, how can one tackle a job that requires indexing for two features that are some specific phase shift appart. Purely theoretical, a discussion on techniques. So, I would like to insist and ask for insight on the specific points/examples:
1. Please have a look at the draft sketch I attached a couple of posts above. Let's assume that the part requires milling slots, as seen in the sketch, each spanning exactly, lets say 12 10' 32" degrees. I understand how to make the start of the three slots, divide by three on a rotab. I also understand how to make the one slot as well: remove the dividing plates, install the degree wheel, zero and count your angle. But I cannot understand how to make the other two ones UNLESS one combines a dividing tool with a degree-based rotating tool.
2. Please have a look at this example, the external lap holder by American Laps
http://www.americanlap.com/External Laps.htm . This can serve as an example to make my second point. Let's assume that one needs to make this part specifically drill the holes. Let's assume also, as an example only, that some features of the part (e.g. the slot) exist before our indexing operation, so our initial setup MUST take into account these existing features. What would make sense to me is mount the part in the indexing device on the chuck, lock the indexer, have a way to rotate the part ACCURATELY in order to locate it as desires with respect to the existing features, lock the rotation and start indexing in order to drill. Again, an independent way to introduce a phase shift of the spindle with respect to the indexer is required. (part shown as an exersise/example only, of course there is no need for such accuracy here again).
3. In our actual part, again as an exersise, assume that the sides of the ribs needed to be radial (imagine it blown up and the ribs being huge, something like this
). In this case, of course, the latteral translation method that Derek/Ross described will not provide the desired result, there has to be rotation in order to mill the other side of the slot. In fact, due to the tool not being infinetely small in diameter there has to be BOTH rotation (by theta, in order to brind slot side 2 where side 1 was) and lateral translation of the tool (equal to its diameter, from location '1' to location '2') in order to make both sides of the slot radial.
Please consider the above examples and, if you have the patience to invest some thought, please provide some insight.
I need to restate here that this is a theoretical discussion, discussion of principles. We all cut corners all day, but the best way to cut corners is to be familiar with the proper way to do stuff. It is better to own the right method and simplify on this as per the requirements than being familiar only with approximate methods that cannot be enhanced in accuracy when it will be needed to do so.
Also, being kind of familiar with the specs of the Deckel machines (and the obsession of their users...
) and the multitude of attachements that allow precision machining in most occasions I would be surprised if they had not provided a tool to do what I am asking here. Even if you consider the simplest of my examples above, no 1 or 2, I can't imagine Deckel tooling system not having respective provision.
Thanks
Thanos