[Kiwi]

For the pitch to start as 63 and end as 169.5 while rotating 720deg the total length will need to be 295.5mm.
Here is the code in steps of 45deg.
G90 G01 X63 Y0 Z45
X72.2626 A405
X82.0556 A450
X92.4093 A495
X103.3559 A540
X114.9292 A585
X127.1652 A630
X140.1018 A675
X153.7791 A720
X168.2396 A765
X183.528 A810
X199.6918 A855
X216.7812 A900
X234.849 A945
X253.9514 A990
X274.1475 A1035
X295.5 A1080

[Kiwi]

Sorry for bombarding your post with code but.....
I now think there is a change of pitch between 63 and the end.
I believe this is what Geof has been saying it should be like to be a 'Simple Harmonic Motion'
I'm not sure if the sketch is to scale but if there is a change of pitch after the 720deg point a path can be generated much closer to the sketch.
I now have the pitch changes at 198mm (Pitch 207) and at 375mm (Pitch 147)
Here is the XYZ code for you to examine.
G90 G01 X63 Y0 Z45
X66.2617 Y11.6468 Z43.4666
X69.6683 Y22.4999 Z38.9711
X73.2264 Y31.8198 Z31.8198
X76.9426 Y38.9711 Z22.5
X80.8239 Y43.4666 Z11.6468
X84.8778 Y45 Z0
X89.1118 Y43.4666 Z-11.6469
X93.534 Y38.9711 Z-22.5
X98.1528 Y31.8198 Z-31.8199
X102.9768 Y22.4999 Z-38.9712
X108.0152 Y11.6468 Z-43.4667
X113.2776 Y0 Z-45
X118.7738 Y-11.6469 Z-43.4667
X124.5143 Y-22.5001 Z-38.9712
X130.51 Y-31.8199 Z-31.8199
X136.7721 Y-38.9712 Z-22.5001
X143.3125 Y-43.4667 Z-11.6469
X150.1437 Y-45 Z-0.0001
X157.2784 Y-43.4667 Z11.6468
X164.7302 Y-38.9712 Z22.5
X172.5133 Y-31.8199 Z31.8198
X180.6422 Y-22.5001 Z38.9711
X189.1324 Y-11.6469 Z43.4666
X198 Y-0.0001 Z45
X206.6438 Y11.6468 Z43.4666
X215.1655 Y22.4999 Z38.9711
X223.5668 Y31.8198 Z31.8198
X231.8495 Y38.9711 Z22.5
X240.0152 Y43.4666 Z11.6468
X248.0655 Y45 Z0
X256.0021 Y43.4666 Z-11.6469
X263.8267 Y38.9711 Z-22.5
X271.5407 Y31.8198 Z-31.8199
X279.1458 Y22.4999 Z-38.9712
X286.6434 Y11.6468 Z-43.4667
X294.0352 Y0 Z-45
X301.3225 Y-11.6469 Z-43.4667
X308.507 Y-22.5001 Z-38.9712
X315.5899 Y-31.8199 Z-31.8199
X322.5728 Y-38.9712 Z-22.5001
X329.4571 Y-43.4667 Z-11.6469
X336.2441 Y-45 Z-0.0001
X342.9353 Y-43.4667 Z11.6468
X349.532 Y-38.9712 Z22.5
X356.0355 Y-31.8199 Z31.8198
X362.4471 Y-22.5001 Z38.9711
X368.7682 Y-11.6469 Z43.4666
X375 Y-0.0001 Z45

[Geof]

The way I see it is that the first 63mm is a constant pitch, ie 63mm over 360 deg. then the acceleration begins and lasts for a further 312mm.........I will double my efforts to get some clarification from the customer.

Thanks Geof.

Yes I see it now. I was stuck on the idea that the easiest way to get from a pitch of 63 to 169.5 would be to take the average of the two and split it up over the distance.

More clarification is needed I think.

[pbd1971]

Ok guys, the plot thickens Had a look at the original drawing today, its dated 1986! and the guy who did the drawing has retired. The head of design at the place is on his holidays at the moment so it might be a week or so before i can get any clarification on this.

I will upload the original drawing some time over the weekend, might have to be in 2 parts, maybe you can spot something i have missed.

Thanks and have a nice weekend!

[Geof]

Ok guys, the plot thickens Had a look at the original drawing today, its dated 1986!.....

I suspected as much. And I will treat you to a few beers if the current head of design can figure it out any better than you can.

[pbd1971]

Here is the full drawing, sorry its in 2 parts, I only have a basic scanner here.

@Kiwi, you know I am beginning to wonder if I have misunderstood the drawing. I assumed the 169.5 pitch was a representation of the pitch at the end of the screw but it could also be interpreted as stating the pitch of the next part of the production line. Not sure.

Also, I do not have a clue what is meant by the description under the Calculation Header, "Dimension A= Can OD (raised to nearest mm) and Dia B=A+5mm..............There are no A or B dimensions anywhere on the drawing and I cant see how it could be linked to the harmonic calculations

Thanks

[Kiwi]

OK, I finally got a combination that may be what you need.
IMHO. A change of pitch to P195.75 at 192.375mm will finish with P169.5 at 375mm.
Unfortunately the drawing appears to be out of scale but the pic below has this path overlaid.

[HuFlungDung]

Simple Harmonic acceleration is kind of an open-ended specification. A Scotch yoke has 4 phases of harmonic acceleration/deceleration because it operates through a 360 degree crankshaft, but in this screw application, they would only be interested in a sinusoidal acceleration from Velocity A to Velocity B (represented by a 90 degree rotation of a crankshaft). That is, the can coming into the screw has a velocity presumably from a linear conveyor and they want to ramp it up to a new speed when it exits this screw.

A continuously changing pitch of the screw helix is representative of a continously changing velocity (and a continously changing velocity is by definition an acceleration), IMO, and is SHM, certainly good enough for this application, I would say.

[Kiwi]

...I assumed the 169.5 pitch was a representation of the pitch at the end of the screw but it could also be interpreted as stating the pitch of the next part of the production line. Not sure...

The drawing appears to shows the 169.5 pitch beyond the screw, but IMO I would also interpret this as the pitch at this end of the screw as well.

[Kiwi]

Here is another way which may be worth considering.
See pics.
When drawn on the flat, one 360deg path is at 63 pitch, one 360deg at 169.5 and the other 360deg is an arc which starts approx parallel with the 63 pitch and ends parallel with the 169.5 pitch. When rolled looks quite presentable.
A program could be written to extract the X points at the degree increments.

[Geof]

Here is another way which may be worth considering....

.....A program could be written to extract the X points at the degree increments.

Well worth considering. You have come up with a much better approach than I could think of.

Could you extract the X points converted to inches, scale them by 0.65 and put a 'B' between every line? Then I will stick another length of round bar in my machine and let it run for a while.

I need inches because that is what my machine works in.

The 0.65 scales it so the O.D. is about 2-5/16" which is my bar size. It also scales the total length so it will fit within my machine limits.

The 'B' is so I can do a find/replace in my editor and put in a M97 P1000 subroutine call after every X move. This way I can use my interpolation subroutine to cut the 68mm diameter, which becomes 1.74" after scaling, with a 1/2" cutter.

[Kiwi]

...Could you extract the X points converted to inches, scale them by 0.65 and put a 'B' between every line? ....

Geof. Is this the format that you require. This is in steps of 30 degrees.
Do you also want the 63 and 169.5 parts included and same format.
Also what degree steps are you wanting?

X1.6122 A360
B
X1.8171 A390
B
X2.0382 A420
B
X2.2757 A450
B
X2.5303 A480
B
X2.8024 A510
B
X3.0927 A540
B
X3.4018 A570
B
X3.7304 A600
B
X4.0795 A630
B
X4.4498 A660
B
X4.8425 A690
B
X5.2589 A720
B

[Geof]

Kiwi; That is it exactly the format. I take that and replace the B with the subroutine call, and the program I used for my sample in the other thread becomes this:

%
O00004 (BOTTLE SCREW)
(G54 X AT END OF SCREW)
(G54 Y AT CENTERLINE OF PART)
(BAR OD 2.316)
(Z OFFSET TO OD OF PART)
G10 L12 G90 P1 R0.5
T1 M06
M03 S8000
G43 H01
G54 G90 G00 X0. Y-2. Z1. A0.
Z0. M08
G01 Z-1.2 F50.
X1.6122 A360
M97 P1000
X1.8171 A390
M97 P1000
X2.0382 A420
M97 P1000
X2.2757 A450
M97 P1000
X2.5303 A480
M97 P1000
X2.8024 A510
M97 P1000
X3.0927 A540
M97 P1000
X3.4018 A570
M97 P1000
X3.7304 A600
M97 P1000
X4.0795 A630
M97 P1000
X4.4498 A660
M97 P1000
X4.8425 A690
M97 P1000
X5.2589 A720
M97 P1000
G53 G00 Z0.
M30
(-------)
N1000 G41 D01 G01 X0.45 F80.
G03 I-0.45 J0. X-0.9 Y0.
G40 G00 X0.45
G41 D01 G01 X0.87 F80.
G03 I-0.87 J0. X-1.74 Y0.
G40 G00 X0.87
M99
(----)
%

With the coordinates for all sections in one degree increments I can make a very close replica. The faceting at one degree will be about 0.01" at the periphery of the screw.

If the machining time is tolerable I will do two and send you one as a conversation piece.

[Kiwi]

If the machining time is tolerable I will do two and send you one as a conversation piece.

Thanks for the offer, better see how things turn out first.
I saw your other code but didn't really understand if. I have a machine centre with a Fagor control so it's bit different to a HAAS.
I also don't have a forth axis, so I can't do this sort of job. Find generating the code interesting.
Attached is the code I believe is correct. Better do a simulation first.

[Geof]

It is going to work (I am pretty sure).......But Haas machines need a decimal on the A value. A1 is interpreted as A0.001, I need A1. for it to be read as A1.

However, right now as I type, I have my simulator running the Haas Graphics program and it is working. It has reached X2.4873 but only A0.479 not A479.0

My offer to send you a sample was not entirely flippant. I am pretty confident this will work and I will be down in NZ next January so I don't have to pay postage from Canada, I will carry it with me.


EDIT: My simulator tells me the machining time will be 35 minutes, maybe a bit more when the correct A moves are in the program..

[Kiwi]

....But Haas machines need a decimal on the A value. A1 is interpreted as A0.001, I need A1. for it to be read as A1.

Updated version attached. This needing a decimal point seams a bit quirky? My Fagor will handle that format OK.

[Geof]

Thank you.

Yes Haas machines do have some quirks compared to other machines but one quirk I like is the ability to have a local subprogram and call it with an L count. Also call it from a line that includes an incremental move which is how my other program worked.

The program based on your coordinates just calls the subprogram that interpolates the groove after every X, A move. The same thing could be done on a machine that had the subprogram in an independent program but it would probably be a lot slower.

The new program with A. values is currently running in my simulator to get the cycle time. It will take a few days for my machine to be free so I can play on it. Sometimes the real world intrudes and I have to do a project for the company.

[Kiwi]

Geof. I now believe that a smoother path can be machined. The change from P63 to P169.5 over an arc starts at 360deg and now finishes at 535.1deg/118.4mm. I have rearranged the layout to create the path. See attached pic.
I will try and generate some new code so just sit on what you have. Cheers.

[Kiwi]

Geof. ....I will try and generate some new code so just sit on what you have. Cheers.

Modified code attached. When viewed as X and Y code the path look nice and smooth.
I hope this is what pbd1971 really needs.

[Geof]

This one doesn't look so much like a candle holder but it does look very similar to pbd1971's drawings.

Programming time once I had Kiwi's coordinates was about 10 minutes because I had the other program to convert. Cycle time was 25 minutes. The finish is not very good because I was using an old dull 5/8" three flute cutter.

This was a fun project and now my brain is rejuvenated to think about more serious things.

Program is in the text file.