milling crankshaft on cnc mill using A axis

checked all settings bit of a blonde moment when setting the y axis used edge finder as 4 mm dia should be 5 mm so .5 mm off center now no fillets but it looks to me that the journal has a slight rad ?? any help please

Geoff

Hi Geoff, crankpin will be barrel shaped due to your milling cutter being hollow ground by a couple of degrees, you could grind milling cutter truly flat across the bottom or I think you can do another pass around the journals but with Y axis offset by just under half the cutter diameter, this should just flatten off the high spots.

Now you are seeing the cone shape because the end cutting mill leaves a surface higher in the centre. You could get the journal truely cylindrical with an infinite number of passes, each offset by an infinitely small ammount. You will only get the journals parallel if the tool has a flat end. The end product is a roughed out crankshaft ready to be finished turned , or ground.

Even with a flat bottomed cutter you don't get an exactly cylindrical pin. Assuming a finite number of flutes on the cutter you end up with a series of facets that are on a helix. On a practical note the centre of the cutter will be operating at a very small surface speed, so it is likely that the surface finish will vary across the pin.

Andrew

Time to lay the 4th axis on it's back and mount a grinding wheel in the spindle and then alter the various axis in the program so that the wheel will follow the pin's axis as it rotates and give that final ground finish. Should be OK on a short single throw crank like this but would be a bit flexible for a longer multi throw one unless some form of end support for the crankshaft could be arranged

Edited By JasonB on 04/02/2020 20:26:41

After my earlier posts in this thread I read the earlier thread referred to on 3 Mar by Jason and was very surprised to see that it was claimed by John Stevenson that using G19 got complicated quite quickly. I have written a new parameterised program that uses rotation in the yz plane (G19) that should give a better finish than the 360 facet version that John used. I have also added two polishing passes at y=1/4 and 1/2 of the cutter diameter off the a axis centreline to reduce barrelling of the pin. I have given it a physical run and it works fine on my machine. I will set something up tomorrow to photograph and video.

I have not tried to make it efficient so it still does a lot of air cutting but it does not rotate the a axis back to zero with each pass.

The code is posted in the next post due to character limitations, if spacing looks odd it was done in a text editor with fixed character width.

If yo want to see the toolpath in Mach 3 you need to drag the window around with the left mouse button as the default view shows nothing.

Martin C

; This code assumes the units are mm, may work with inches but not checked.
; It is essential that the cutter is long enough to pass between the webs when the pin is at its lowest point.
; y zero and z zero (bottom of cutter) should be on the a axis centre line.
; Set x axis to suit your requirements for each pin.

M0 ; set a axis to zero if required to avoid long wait.

; The following parameters should be set by the user to suit the part being made.

#1=[10] ; Bar diameter
#2=[6] ; PCD of the crank pin
#3=[3] ; Crank pin diameter
#4=[6] ; Cutter diameter
#8=[0] ; Pin offset from TDC.
#9=[1] ; a axis rotation. Set this to -1/+1 if the a axis rotation is wrong direction. Check before cutting any material.
#12=[250] ; Set the speed to suit your cutter and material.
; The following are calculated parameters.

#6=[[#1 /2] + #2] ; Calculated z start height for first pass.
#7=[[#2 /2] + [#3 /2]] ; Calculated z finish height for final pass.
#5=[[#6 - #7] / #2] ; Calculated step size.

; The number of passes is taken as the pin pcd #2.

; This gives a step size close to 1mm but increases if the bar diameter is much larger than the pcd.
f[#1]
#10=[#2] ; Step counter for looping the sub-routine.
; Used as a multiplier to set the z axis before each rotation.
#11=[1] ; Adds 360 degrees to the angle each rev to save rotating backwards each pass.

; Initialisation
G0 G49 G40.1 G64 G17 G80 G50 G90 G98
M6 T1
M03 S4000
G43

G19 ; Set circular motion to the yz plane.

; Start motion

G0 z[#6 + 2] ; Go to initial start height.
G0 a[#9*#8] ; Move a axis to TDC plus offset.
M98 P003 Q#2 ; Loop subroutine #2 times

; Polish passes.

G1 z[#7 + #5*#10] y[#4 /-4] F[#12] ; Reposition cutter to clean up more of the pin.
G03 a[[#9 *#11 *360] + [#9 *#8]] y[#4 /-4] k[#2 /-2] ; Circular motion 360 degrees.

#11=[#11 +1] ; Increase degrees multiplier for final rotation.

G1 z[#7 + #5*#10] y[#4 /-2] ; Reposition cutter to clean up the centre of the pin.
G03 a[[#9 *#11 * 360] + [#9 *#8]] y[#4 /-2] k[#2 /-2] ; Circular motion 360 degrees.

G0 z[#6 + 2] ; Go to initial start height.
G17 ; Set circular motion back to the xy plane.
M5 M9 ; Turn off spindle, coolant etc as required.
M30 ; End program

O003 ; Circular motion subroutine.

G1 y[#4 *-1.5 *#9] F[#12] ; Move slightly to one side.
#10=[#10 - 1] ; Multiplier for z height.
G1 z[#7 + #5*#10] y0 f[#12 /4] ; Ramp down to inital cutting position for circular motion.
G03 a[[#9 *#11 *360] + [#9 *#8]] y0 k[#2 /-2] F[#12] ; Circular motion in yz plane 360 degrees.
#11=[#11 +1] ; Increase degrees multiplier for next rotation.

M99 ; End subroutine.

Had to delete spaces to post it.

Martin C

Not having a 4th axis I look forward to the video but may see what it looks like on the Mach3 screen.

Here are the test pieces from my Gcode. They are machined from a slug from a Rotabroach cutter which I saved for uses such as this and are 20mm diameter (roughly) and 25mm long.

Pins are 5mm diameter on a PCD of 14mm. Offset from one to the other is 120 degrees. Cutter was 1/4" 2 flute.

The pause from one pin to the next in the video is due having to edit the code for the offset of 120 then manually move the x axis then reset the a axis to zero to avoid a long reverse rotation to zero plus offset before running the code again.

Martin C

Edited to embed video instead of a link.

Edited By Martin Connelly on 05/02/2020 15:08:59

Thanks for that Martin, certainly seems to do the job and the machine sounds a lot happier than Geoff's one. I like the small movement in Y each time the depth of cut is increased which saves the cutter plunging into the work and the two finishing passes look to leave a less barrel shaped pin. Easy to see the shape that the end of the cutter leaves at about 3.10 into the video where it makes the first pass on the second pin.

I was thinking that another way the barrel shaped pin could be reduced is to use a cutter that has a diameter less than the gap between webs say a 6mm cutter for a 10mm gap which would need two runs through with a 4mm movement in X for each pin.

During the early 2000's when I was involved in various aspects of engine design I visited a a crankshaft manufacturer in the UK midlands. They specialised in one off and very low volumes, the shafts were typically "car" size. I was surprised to see the method of machining.

The cranks were roughed out from a suitable diameter piece of steel on very heavily built lathes. No top slide, tool post mounted directly on the cross slide in the "front" position. The tool looked like a very substantial parting tool with a screw jack arrangement supporting the tip of the tool. Even with one machine in operation the building seemed to shudder every time the lathe took a cut. I found this very wearing, frankly glad to get out of the building after a few hours of constant noise and vibration!

The roughed out cranks were then transferred to 4 axis CNC machines that finished the cranks prior to heat treatment, grinding, radius rolling and peening/polishing. Watching the speed and ease at which the CNC's worked I asked why not rough out on the CNC as well? Apparently quicker to do it as they were, was the reply. I wonder if they still do.

John

I am going to try Martins program to see how well it works compared to the one by JS, I have had the JS one working well after I changed the rotation of the a axis.

I am a bit thick most times so would like to know what "pin offset from TDC " means please.

regards

Mark

If you do two pins at top dead centre (Zero offset from TDC) they would be on opposite sides of the web and both at TDC together. In a four stroke four cylinder engine there are usually 4 pins with 1 and 4 at top dead centre the others on bottom dead centre (2 and 3). In this case when cutting the first pin do it at top dead centre. The second and third pins would have their offsets from TDC at 180 degrees then pin 4 would be back to an offset of zero from TDC. For three cylinder engine you could have 1 pin at TDC and the other two with offsets of 120 degrees and 240 degrees from TDC.

Hope this explains it, in the video I did the second pin has an offset of 120 degrees from TDC. The pause between finishing the first pin and starting in the second one is where I manually changed the X axis position and then changed the offset to 120 degrees before starting the program. You could easily edit the program with cut and paste to set this up automatically (ie no manual intervention) for a multi cylinder crankshaft.

Martin C

Martin,

thanks for the reply! I do understand that bit now! Not sure if I am able to edit the program to auto move it to the next crankpin it all gets beyond me! Will try and do a Seal 15cc crank in aluminium to manually moving to the next crankpin and see how it goes. I have a V8 crank I want to do that would be nice if I can cnc it

thanks

Mark

Interesting... Just making my first steps into CNC milling