Dynamic milling vs pocket and distorsion

[jslaero99]

Good afternoon, I'm wondering if anyone out there made a test piece to compare distorsion results between standard 'high speed' pocket' vs. dynamic? Let me explain the situation, we are 10 programmers here, 5 on Mastercam and 5 on Catia. We were making parts with Catia using standard pockets (±.300 doc) with good flatness results (within the .004 required). Mastercam programmers started doing similar parts on the same machine using dynamic (±.800doc) and the flatness is out of tolerance. We looked at some parameters but I'm starting to dout about the stress generated by the dynamic in structural aluminium parts. We should be doing a test tomorrow morning but wanted to hear from you guys.

 

Thanks in advance!

 

JS

[djstedman]

Are Both parts the same overall depth, and what is the overall depth?

 

From your explanation it sounds like the 'standard pocket' one is being stepped down .3 each step and the dynamic mill is stepping down .8? Am I understanding that correctly?

 

Anyhow.. I havent tried doing a side by side comparision but it seems to me it would be very dependent on the geometry of the part, fixturing, etc .. all things being equal I would think it would depend on the amount of force beign generated by each toolpath, this could change dramatically based on width of cut in the dynamic mill path as much as in the standard pocketing toolpath.

 

Just thinking about it I would think a path .3 deep and .5 wide would create a lot more force than a path with a constant depth of cut of .8 and a width of cut of .05 - .075

[jslaero99]

You do understand cutting parameters correctly. Parts are very similar at ±2.5 inches deep mounted on the same (horizontal) machine, both picture frame fixtured with invert-a-bolts. Your theory sounds good, because the dynamic is .8 doc and .3375 (45% radial engagament) and the pocket is ±.3 x .5 rad. eng. So in this specific case, the dynamic would genenrate more force. I'll keep you posted on the results.

 

Thanks again!

 

JS

[rogkick]

More force, more thermal build-up and transfer, and hence the distortion I would have thought.

[gcode]

and the flatness is out of tolerance

 

the whole part or the pocket floor?

I would never use a dynamic path to finish a floor.. I'd rough with the dynamic then do a finish pass on the floors and walls

with a conventional toolpath

[Sticky]

the whole part or the pocket floor?

I would never use a dynamic path to finish a floor.. I'd rough with the dynamic then do a finish pass on the floors and walls

with a conventional toolpath

 

+1 you are going to have a lot of radial deflection with the large axial engagments used in dynamic/high speed milling.

 

I use the same technique gcode mentions for pockets, don't have any flatness issues this way.

[MotorCityMinion]

Same machines, different tool paths. IMO, it's has to be how the machine is reacting to the tool path. The cutter can flex with adaptive style tool paths, small scallops ensue. The worst condition I've seen with this on a beat up, sloppy Haas is about .0015 dig. The floors still come out relatively flat, just not where I want them. Come to think of it, this worn out machine will do this with any aggressive tool path. The material may also be curling up throughout the roughing routine as well. 45% radial engagement is a big chunk.

 

???.004 ??? Something else is at work here. Although I've never had a problem with the tool path actually being wavy, save the tool path as geometry in back plot and measure some depths.

 

"I would never use a dynamic path to finish a floor.. I'd rough with the dynamic then do a finish pass on the floors and walls"

 

I agree. It's a roughing routine. Spring the part after roughing and re-cut anything critical.

[Matthew Hajicek - Singularity]

How thick is the floor?

[mkd]

???.004 ??? Something else is at work here.

Spring the part after roughing and re-cut anything critical.

^^yup!!^^

is one program producing steam from water based coolant?

steam = heat.

heat = distortion.

 

small radial stepovers in mastercam should be a benefit over old school-gouge-it-out paths.

although, greater axial engagment can pull up more due to the helix angle of endmill.

go with a lower-helix angle cutter?

CORNCOBB rougher.!! great easy fix that usually makes a huge difference in aluminum roughing where flatness is a concern

[Newbeeee™]

CORNCOBB rougher.!! great easy fix that usually makes a huge difference in aluminum roughing where flatness is a concern

We use the MA Ford 3 fluts. Work like a charm.

I wouldn't use 45% stepover with dynamic paths - too high IMHO. I'd go deeper but less stepover - peeeeeeeel it off

[Chris Rizzo]

Disclaimer: I am not a tooling or material engineer. This is just a shade-tree theory of metal cutting. (and a bit of trials)

 

 

wouldn't use 45% stepover with dynamic paths - too high IMHO. I'd go deeper but less stepover - peeeeeeeel it off

 

I would agree. That does seem pretty heavy for dynamic. . That 45% stepover might be putting some heat in, you've got a lot of tool engaged in material. Tool in material=heat....with a more conventional 10-20%, you have 80-90% of the tool out of the cut, thus being in a cooling phase. (I attribute that theory also to being able to run sfm that is off the charts).

 

Go with less stepover and faster feeds and/or more doc. I'll bet you can get your mrr up with . I recall reading a very technical article a few years back regarding aerospace parts (deep thin walls) and this type of situation. I'll see if I can dig it up. I'm curious now.

 

I'm thinking there are some mechanical stresses being relived in different way via the different cutting methods. Also maybe tool maybe sucking up floor? How thin?

 

Totally unrelated to your problem, but what is the cycle time comparatively?

 

never use a dynamic path to finish a floor

 

Here is a tip to finish floors with dynamic:

Run your regular roughing dynamic however you like (usually the 10-15% stepover) Leave .010-.020 on the floor.

COPY the toolpath and kill any depth cuts. set the stepover to something quicker, 40-60%. Oh, and reduce your feed if your roughing at 600ipm

 

At the college I've been running alot of tests with sandvik tooling. Cutting 1018 at 600ipm. .500 cutter, 15% stepover. 1" doc. Their engineers said we'd get 15min of tool life. I'm on 80min right now and still cutting! Working on a video of it. Stay tuned.

[jeff]

 

Here is a tip to finish floors with dynamic:

Run your regular roughing dynamic however you like (usually the 10-15% stepover) Leave .010-.020 on the floor.

COPY the toolpath and kill any depth cuts. set the stepover to something quicker, 40-60%. Oh, and reduce your feed if your roughing at 600ipm

 

 

This is what I do, quick n easy.

[Josh Kinney -Elite Engineering]

Yup. And I only use "micro lift" when doing floor finish passes.

[jslaero99]

Thanks a lot guys for the toughts, MKD, any kind for the CORNCOBB? Matthew, the floors are varying from .08 to .200. GCODE, datum -A- (where flatness is required) is a truncated flat surface at the top of these pockets and cut-outs and BTW we only use the dynamic to rough. Motorcity, the finish ain't wavy as well as the toolpaths it's the part is machined picture frame and the part twists. We're not far .006 for .004 but as mentionned before with standard pockets in Catia, we were more in the .001-.002. We'll make a test with dynamic with lighter DOC. I'll keep you posted.

 

Thanks again!

[kccadcam]

You do understand cutting parameters correctly. Parts are very similar at ±2.5 inches deep mounted on the same (horizontal) machine, both picture frame fixtured with invert-a-bolts. Your theory sounds good, because the dynamic is .8 doc and .3375 (45% radial engagament) and the pocket is ±.3 x .5 rad. eng. So in this specific case, the dynamic would genenrate more force. I'll keep you posted on the results.

 

Thanks again!

 

JS

That heavy of a radial engagement will absolutely cause more stress in the material

With Dynamic milling the radial engagement should not be more than 25% of the tool diameter...

This is because of chip thinning, think light radial load at ballistic speeds.

 

 

 

Read here:

http://www.volumill....ite_Paper_1.pdf

 

http://cuttingtoolsc.../chip-thinning/

 

http://www.cuttingto...hipThinning.pdf

[Colin Gilchrist]

Thanks a lot guys for the toughts, MKD, any kind for the CORNCOBB? Matthew, the floors are varying from .08 to .200. GCODE, datum -A- (where flatness is required) is a truncated flat surface at the top of these pockets and cut-outs and BTW we only use the dynamic to rough. Motorcity, the finish ain't wavy as well as the toolpaths it's the part is machined picture frame and the part twists. We're not far .006 for .004 but as mentionned before with standard pockets in Catia, we were more in the .001-.002. We'll make a test with dynamic with lighter DOC. I'll keep you posted.

 

Thanks again!

 

 

When you do your test, don't change the Depth of Cut, change the Radial Engagement. My general starting point is 20-30% radial engagement in Aluminum, 10-15% in Steels, 5-10% in HRSA/Titanium/Stainless. (*At a 1:1 Depth-to-Diameter ratio)

 

Try to maximize your DOC. That is where these paths really shine (and increase the MRR without increasing the wear/load on the tool). The toolpath does this by never exceeding the radial engagement you set in your parameters. I really like to try and maximize using most of the flute length on the cutter when roughing. If your DOC is programmed at .3, then you are just wearing out the tip of the cutter, and never really use the extra flute length of the tool.

 

The place I like to start when figuring out a dynamic path is looking at the tool parameters and the total cutting depth. I see you are using a .750 endmill to rough this pocket (based on your radial engagement). What is the flute length of this tool? Can you go 1.500 DOC? If so, here is what I would recommend:

 

Total pocket depth: 2.5

Stock to leave on Floor: .05

DOC: 1.225 ( 2 DOC passes)

Radial Engagement: 15%

 

With a Depth-to-Diameter ratio closer to 2:1, you generally want to reduce the radial engagement from the starting values I mentioned above.

 

Even at the increased DOC, the 15% (.1125) radial stepover will significantly reduce the cutting force required. (The cross section of 1.225 DOC x .1125 RE = .1378 sq/in)

 

The old dynamic parameters: .8 DOC x .3375 RE = .27 sq/in

Your Catia Pocket: .3 DOC x .3375 RE = .10125 sq/in

 

*Assuming the radial engagment of the pocket in Catia was the same*

 

Hope that helps,

 

Colin