MIKE@RYTHAN: Usually, X and Y determines R, so most files are X,Y or A,R (angle and radius). Maybe you could come up with more information as to what X,Y and R represent.

Todd: I have a special program to do this. If you send me your whole offset file, I will run it for you. I'll send you the program anyway, but I want to make sure my program runs with your data. I'm assuming a flat cam follower. If it's a roller follower, you'll have to tell me. Basically, the program intersects a bunch of lines which represents the follower as it rotates about the stationary cam.

To avoid math, you could lay out the helix angle in the flat, and roll it on a 1.682 diameter. In the flat, the line for the helix should be at zero depth. When rolled, it will be a helix at the .841 radius.

If you want to do the math, find the angle in degrees = 4 + 22/60 degrees = 4.3667 deg.

The lead, then, is tan(4.3667) * pi * 1.682 = .4035

You could create edge curves and unroll them, or create slices and unroll them.

Sorry, skip the extra 1.

fbuf 1 1 6 0 0

One thing you can do is set the 'keep' bit:

fbuf 1 1 1 6 0 0

Then you can look at the buffer after a run and see what you have acutally put in it.

This is for smtc09. For straight bevel gears, the end tooth form is close to a spur gear tooth form, but not exact, I think. Anyway, all the tooth elements (straight lines) converge at a point, so a fixed tilt on the table isn't going to help that much. I would choose the profile to be perpendicular to the pitch line.

Parameter 4 is the 'read' flag, which means that the file already exists as an input, and you should find the size of it.

size1=rbuf(1,0)

The zero index, or read count triggers the size read, since the read count for reading data starts at 1.

MH says that the term "serrations" was dropped in the 1970 standard, which I take to mean that all inch splines are involute now in the ANSI B92.1-1970 standard.

Machinery's Handbook and Ash catalog. Most of Ash tables appear to come from MH, except for the tolerance table on G-84, which doesn't appear in MH.

Here's both External and Internal dimensions.

SPLINE PARAMETERS

10 NT Number of teeth INPUT PARAMETERS

32 DP Diametral pitch

5 FIT Fit 1..5

7 CLASS Tolerance class 4..7

45. PA Pressure angle

0.22097 DBASE Base diameter

0.31250 PD Pitch dia. (REF.)

0.05534 SWBSC Minimum effective space width

0.05778 SWMIN Minimum actual space width

0.06012 SWMAX Maximum actual space width

.00234 TOLM Machining tolerance

.00244 TOLV Variation allowance

.09817 CP Nominal circular pitch

.04909 HCP Half circular pitch

INTERNAL SPLINE PARAMETERS for GEAR.DLL

0.35625 DMAJI Outside diameter

0.29375 DMINI Root diameter

0.28975 DFORMI Form diameter, internal

0.06000 DPINI Pin diameter

0.03806 TTMINI Circular tooth thickness (for max. space width)

0.04040 TTMAXI Circular tooth thickness (for min. space width)

0.06012 SWMAX Max. actual space width

0.05778 SWMIN Min. actual space width

0.22655 MIMAX Max. roll measurement between pins for max. space width

0.22393 MIMIN Min. roll measurement between pins for min. space width

EXTERNAL SPLINE PARAMETERS for GEAR.DLL

0.34375 DMAJE Outside diameter

0.28125 DMINE Root diameter

0.27725 DFORME Form diameter, external

0.06000 DPINE Pin diameter

0.03806 TTMINE Circular tooth thickness (for min. tooth thickness)

0.04040 TTMAXE Circular tooth thickness (for max. tooth thickness)

0.39859 MEMAX Max. roll measurement over pins for max. tooth thickness

0.39640 MEMIN Min. roll measurement over pins for min. tooth thickness

Sorry. Somehow I drifted into 16/32 mode instead of 32/64. You are right with the pitch diameter. OD is .344, root diameter is .281. Picasso is gone.

The numerator (32) controls the pitch diameter, and the denominator (64) controls the tooth depth, which is half the depth of a gear with the same pitch. The Ash catalog shows a major diameter of .724, and a minor diameter of .588.

Use the diametral pitch = 32 in GEAR.DLL, and key in the OD and root diameters.

MIG, If you can send me the same part file and post that you're using, I might be able to straighten this out. It might help Mike-T too.

Sorry. These parameters, CW, -90, are in the ROLL/UNROLL function. If other paramerers are used, unrolled geometry will appear in the 'wrong place'. If your sample geometry is already flat, the postprocessor may be moving the angle (adding 90 degrees, etc.).

You need to use Rolldie parameters CW -90 to get the output in the right place.

You may be running into the 5-axis 'twilight zone', where the tool axis is nearly parallel to the rotary axis. The rotary axis can rotate unpredictably for a very small xyz move. You might try a lead/lag angle. That would change the relationship between the tool axis and rotary table axis. It could also give you better cutting. It would be easier to figure out if you put your part on the FTP site.

msn_jrd -- I have a surface for you, if you're interested. Just send me an e-mail. I made all the radii of the across arcs the same -- 2 in.

E-mail me if you want the file.

It appears that you want the end 'arcs' to lie on a surface, but you haven't made the surface yet. Trim the large arcs so the surface doesn't come to a point. Just let the large side arcs go a little past the end arcs, say .050. The make a surface and project the planar arcs to that surface. You may want to make 'across' arcs at the ends to finish defining the surface (front view arcs).

The surface is OK as it is. If you want just a part of it, make a trimmed surface. When you 'remake' a surface like this one, you end up distorting it. Trimmed surfaces were invented in the first place to avoid this kind of distortion. A trimmed surface is an original 'good' surface, and a boundary that says 'I only want this much of this surface'.

It looks like the surface is OK as designed, like part of a ruled surface which has all elements parallel. Make a 2d toolpath of the racetrack and project it to the surface.

FPLOT.DLL is a c-hook. When you run the c-hook, there is an option to read the -.eqn file (fig8.eqn). You may wish to edit the diameter and width first, with a text editor. The key to the figure 8 is that one cycle of the sine curve (360 degrees) wraps twice around the diameter. Put FIG8.EQN in the c-hook directory with the other examples.

In general, the math capabilities of FPLOT are supposed to be similar to the postprocessor capabilities, but it falls far short, as there are only two canned loops and no subroutines or other input source, such as a -.NCI file or buffer files. I'll send you both FIG8.EQN and FIG8.PST.

If you want to lay out a sine curve figure 8, here's a FPLOT program that could help. After you plot the sine curve, roll it on the appropriate diameter. You can edit the diameter and width in FPLOT.EQN.

dia = 1

width = 1

step_var1 = a

step_size1 = 5

lower_limit1 = 0

upper_limit1 = 360

geometry = lines

angles = degrees

origin = 0, 0, 0

x = sin(a)*width/2

y = 2*a*pi*dia/360

It's probably either 14.5 or 20 deg. pressure angle. Plot both gears in Mastercam, and see how big a pin you could fit in the root.

Curve5ax has the lead/lag you want. The fillet CL would be the toolpath of the ball center. But CURVE5AX needs a surface to reference for the initial surface normals, to which it adds the lead/lag. Since the fillet surface is the same distance from the CL, you can't use that. Make a ruled surface between the rails of the fillet surface, and use that as your reference surface for CURVE5AX.