NPT tapping

[inertialabs]

I have a project that I am working on and I could use some advise I am making a part out of 303 SS and it has a NPT 1/2 x 14 threads on it one side is female and the other male

I am running the parts on a VF3 SS using Virtual Gibbs and I am using a 60 deg single point cutter and it is working great but it is taking way to long to cut the threads,

With Gibbs it only supports single point threading but I would like to use a NPT threadmill my guess is this will speed up the process by a lot ,

Now to my question

Does Haas have a Canned cycle that will support NPT threadmilling ?
with a muilti point endmill

Can someone tell me if this will be a faster method and produce a quality thread

Thanks

[gizmo_454]

Ah yes, the good ol' single pointer. Been there, done that. They work well, but as you said, S-L-O-W-L-Y. Haas does not have a canned cycle for thread milling, of any kind, NPT or standard.

Somebody pointed me to, I think, Micro100's website for an Excel spreadsheet that does the work for you. You just put in your cutter dia., feed per tooth, # of teeth, sfm, depth of hole, and whether or not you want cutter comp on, and it creates the code for you. I normally use the cutter comp. Set your cutter dia. in the Haas to 0 and start there. I find that usually it gets dialed in to the right size somewhere between -0.005" and -0.015".

If you are interested and cannot locate the spreadsheets online, email me and I will send them to you. I have one for single pass standard threads, one for single pass NPT, and one for 2-pass NPT.

Mike

[Matt@RFR]

You can still use what Gibbs spits out, but hand edit the code and take out all but the bottom full-pitch worth of G2/G3's and let 'er rip. That will get you one pitch of Z feed, and now you can use a full threadmill.

[gizmo_454]

That is true. You can try to do it that way. And it may work. But for NPT threads, it will not work that well. That is, unless Gibbs does it correctly. I've never used Gibbs before, so I don't know.

Because NPT is a TAPERED thread. If you run a true circle as you move vertically and get to the end of the circle, there will be a mismatch. If Gibbs will run a spiral as it is moving vertically, the thread will work fine.

Just my 2 cents.

Mike

[inertialabs]

Originally Posted by gizmo_454 That is true. You can try to do it that way. And it may work. But for NPT threads, it will not work that well. That is, unless Gibbs does it correctly. I've never used Gibbs before, so I don't know. Because NPT is a TAPERED thread. If you run a true circle as you move vertically and get to the end of the circle, there will be a mismatch. If Gibbs will run a spiral as it is moving vertically, the thread will work fine. Just my 2 cents. Mike

That makes sense, so far the single point is working so I can be confident that the NPT thread has the spiral and the code is good but it sounds like I just need to limit it to 2 turns max then come up out of the hole

Thanks for the advise

[inertialabs]

Originally Posted by Matt@RFR You can still use what Gibbs spits out, but hand edit the code and take out all but the bottom full-pitch worth of G2/G3's and let 'er rip. That will get you one pitch of Z feed, and now you can use a full threadmill.

Here we go, I got Gibbs to make a spiral from the bottom up and it goes 1.5 turns up here is the post I got from it but I must admit I am not able to just look at the code and confirm that it is correct I will post it here if you would like to see it

%
O1
G0 G17 G40 G80 G90 G54
T17 M6
S289 M3
G54 G90 G0 X0. Y0.
G43 Z1.1 H17
M8
Z.1
G91
G1 Z-.5625 F1.16
G0 X.0513 Y-.0513
G1 X-.0069 Y-.0041
G2 X-.0444 Y.1729 Z.0325 I-.0444 J.0807
X.0566 Y-.2205 Z.03 J-.1175
X-.0444 Y.173 Z.0325 I-.0444 J.0808
G1 X.008 Y-.0004
G0 X-.0202 Y-.0696
G90 Z.1
M9
G91 G28 Z0.
G49
G90 X0. Y0.
G91 G28 Y0.
G90
M1
T17 M6
M30
%

Thanks for the help