Need Machining Help

[RalphW]

I am using a PMD-30 miling machine in an attempt to make some GPU and NB blocks. The mill can be seen here:
http://www.kingcanada.com/Products.htm?ID=121

The problem is the finish that is left on the base. I have the spindle set to the max. 2750 RPM and am trying to get the flatest, smoothest surface I can using the 3" facemill that came with the machine.
There are slight ridges left on the surface after taking a pass of between .005" to .010"

Would new or different types of inserts in this facemill provide a smoother surface?
Is there a better tool to use than a facemill for this application?

Ralph

[killernoodle]

I dont know, but you may not be able to achieve the face milling accuracy you want with that machine/face mill combination. Hand lapping is probably a good idea.

[RalphW]

You maybe right.
It's just that the hard hand lapping that is required to make the block smooth, also makes the block convex. Perhaps my method is wrong.

What type of facing tool gives the best finish anyway?

Thanks

[killernoodle]

Have you tried milling it in different directions all the same height so the scratches are evened out?

[RalphW]

No, I have not tried that.
I have been milling in just one direction; feeding the material into the cutter.

Will try as you suggested.

Thanks

[killernoodle]

Also, try sloing down the feed rate to nearly nothing and cranking up the rpms (if you arent already maxing out).

[Groth]

Your spindle and table may not be perfectly perpendicular. Feed peices through from the four different directions and compare. If the finish isn't identical, adjust your machine.

You might also try an bull-nosed insert. You won't get the cutting speed of a chisel or pointed insert, but you'll get a smoother finish.

[Incoherent]

I'm no expert but I have had the same problem with a fly-cutter. 2750 RPM sounds awfully fast to me for a 3" diameter cutter. Are the blades blunt?

[Kobuchi]

Originally Posted by Groth

Your spindle and table may not be perfectly perpendicular. Feed peices through from the four different directions and compare. If the finish isn't identical, adjust your machine.

Worth repeating.

That'll be a dynamic calibration too, the best kind.

[JFettig]

you say its concave when you lap it? that and the looks of the finish it looks like the head of the machine is leaning forward a little, It also looks like you are cutting on the Y axis, try using the X axis(left to right)
One way to test and see if its not completely squair is with a dial indicator, chuck up the rods from a magnetic base and swing it from one side of the table to the other, forward and back.
another way to just see if its not squair is cutting with the x axis and cut only part way, then come back and cut more(move Y to cut only half way across, then cut more) and see if it leaves a groove.

I have had these problems on my mini and still kinda do. Its tricky to get it fixed.

-edit- If your mill is completely squair, the finish should look similar to this one with marks on both sides:


Jon

[killernoodle]

I wish the bottom of my maze 4-1 looked that good JFettig...

[RalphW]

Thanks for the suggestions.

I am not going to get a chance to try any of these things until later tonight.
I did put a mag base indicator on the table and measured for any runout on the head colomn (while moving the head up and down) but I only took a reading off of the side of the column and not for front to back tilt.

Will report back with some results.

[ataxy]

ballnose tip will definitly give you a smoother finish
also what cutting depth do you take cuz often people mistakingly think that by taking a small layer cut they will be able to sort of polish there suface

[RalphW]

I did some test today.
First I wanted to check if the table is perpendicular to the spindle. I mounted a dial indicator to the spindle column and measured off the table bed as I ran the x and y axis back and forth. There was less than .002" runout.
Then I placed the mag base on the table and measured off the spindle column as I moved it up and down. I checked from the side and the front and again less that .002"
So I assume I can rule out the machine squareness?

I cut the piece of copper shown in the diagram in all four directions. The swirl marks evened out but some area's look better than others.
The result is better than the one direction cut on the block beside it.

The inserts feel sharp but I think I will replace them with some bullnose one's as was suggested.

Thanks for the help.

[jaydee]

Look like the cutters might be dull to me. See the edges of the right pic? Looks like there is some copper still attached on the edges over hanging. Sharp cutters should clean that up well. Also check for any vibration and make sure the fly cutter itself is not off balance and the cutters are all inserted at the same level.

Good luck, you will get it figured out. Nice setup.

[SlaterSpeed]

First off whatever happens you will get machining marks. The only way to finish your blocks base on a machine is to gind it but unfortunatly copper is too soft to gind easly (and im guessing you dont have a surface grinder to hand)

The alinement of the head wont relly make alot of differance to surface finish aslong as its within normal tolorances (if its not then your going to be cutting a concave base anyway so its still important as ever)

The way you clocked up on the quill is a pretty bad way of doing it. A better way is to get yourself a bit of rod about 6" long, put 90deg bend on each end in oppisite directions. Then attach your indicator on one end and hold the other in the chuck (you can buy something simlar but its easyer to make). Take the mill out of gear and swing the indicator around on the bed 180deg checking the dial stays the same.

Try bull nose inserts but what ever you do your still going to have to lap it

slater..

[RalphW]

The way you clocked up on the quill is a pretty bad way of doing it. A better way is to get yourself a bit of rod about 6" long, put 90deg bend on each end in oppisite directions. Then attach your indicator on one end and hold the other in the chuck (you can buy something simlar but its easyer to make). Take the mill out of gear and swing the indicator around on the bed 180deg checking the dial stays the same.

I could mount the rod that the dial indicator is on directly to a chuck. But why do I want to swing the head back and forth? Can't I just move the x and y axis through their travel to check that the bed is square to the head?

Thanks

[ataxy]

no cuz the idea is to see if the head is tilted if you move the table and not the indicator you wont see a difference since the indicator does not move all you will see is the small inperfection on your table surface but the mesurment will stay the same

[RalphW]

Thanks

I believe I understand the concept now.
Will try that out and see what kind of result I can get.

[Kobuchi]

"Your eyes can deceive you; don't trust them."

Though I've used something like SlaterSpeed's "Z" rod method, I don't like to intervene the measurement with those two fine instruments: the dial gauge, and the human eye. This just adds error upon error. Try grazing the table with the rod itself, and listening. Variations in pressure yield distinct sounds.

Of course this method is hopelessly subjective, but an attentive individual with normal hearing can get nearer true using it than by splitting hairs squinting at a dialface, or worse yet reading numbers off one.