4th axis slots 1 of the 3 are centered?

[CuttingEdgeManufacturing]

I'm stumped on this one.

i have this program i designed, and programmed in F360. we just machined the first one, and 2 of the 3 slots that get machined early in op 3(setup6) are off center. i have no clue how this can be. i've checked all i can check. but the code shows that only 1 slot is cutting at y-.376 one at y-.3888 and one at y-.363. 

to make this program i just used tool orient on 1 slot and then created a circular pattern, using the OD of the part as the axis of rotation.

does anyone have a clue why its doing this? 

[Anonymous]

There is a lot going on there so I am not going to study it fem A to Z but I suspect that in setup 6 you may have used Z axis orientation that is not valid due to potential misalignment of part with fixture.

If you use bottom of 3 wide slots as your Z axis reference instead of flat on fixture, what do you get?

Also in, circular pattern you used segment of circle as axis of rotation, I would prefer to use full circle such as ID of the part or X axis of triad for more accurate alignment because you are using large face of the part as X axis reference.

[Anonymous]

May wanna flip direction of X axis on my last screenshot but I am pretty sure using fixture in setup is a clue to a alignment problem, align axis to features on part, if there are none that serve purpose, sketch lines constrained to axis of triad or model and use them as axis orientation in setup.

[CuttingEdgeManufacturing]

i dont get any errors at all in f360 and simulation of part shows that all is well. i am not referencing the fixture at all except for the original Z axis of the setup. but i have to do that otherwise i cannot locate on A axis. not sure what to do.

[Anonymous]

I had a case where milling on a turned copper part was done on 4th axis, I made all fixtures, programmed entire job from start to finish. Part had .03 wide channel on inside wall and .003 wall thickness between two windows.

In order to do this right, I instructed operator to indicate fixture to as close as possible accuracy, he ended up damaging part of the seat for part which dislocated that channel from what would otherwise be even wall thickness on each side.

So the point I am trying to make is that fixture being slightly out of alignment translates into features on part being out of alignment, Fusion simulation will not show that unless you deliberately miss align fixture with model and measure the result.

So I think you may have an issue with fixture not being properly indicated which would for sure result in features being out of sink as A axis is rotated to angle of each milled features.

Something gets lost in translation between Fusion and machine, the other possibility is that perhaps features on your model are not accurate, Fusion will simulate tool paths but won't tell you that 2 slots are out of line with third by some mediocre value.

[CuttingEdgeManufacturing]

hmmm. as far as im understanding is that the problem originates in F360. i say this because if i post the code and look at it, the finish pass for the 3 slots have different y values.

but that doesnt make sense, because i programmed 1 slot and simply rotated it around the part. i have checked my model over and over and as far as i see everything is concentric.  BUT my program doesnt use the model to cut all 3 slots. it uses 1 slot and revolves the toolpath about 3 positions. 

im seriously considering this one to be a flute or error in fusion. i cant seem to see what else it could be.

on another note, i did have to change the rotation of the axis in the post processor for some reason on this part. it tried to machine my spotfaces for the reamed holes in the wrong direction.

[Anonymous]

I went back to your file looking for any misalignment between model and fixture and couldn't find any.

So if you narrow it down to Fusion, then looking at each setup, you are using faces as perpendicular reference for axis, I don't see any diameter being used as reference feature in setup that would ensure concentricity of milled features to model longest section like ID.

Setup 5 has nothing referencing X axis, if I consider the fact that you had to change something in post, I am lost as to why,... did post fail to produce correct code because of missing or inadequate axis alignment with model ?, I don't know.

Does using another post yield same result ?

Back to machine,.....

If you set indicator near outer edge and on large face of model, does it read zero when you manually rotate X axis?

If you sweep across X axis over model ID wall from end to end, does indicator needle stay at zero?

Is work offset set precisely in the center point of model?

May be redundant but I would suggest using different features for axis reference in setups and posting with another post to see if there is a different outcome.

I would also double check machine setup in hope to uncover potential error in work offset and alignment. 

[CuttingEdgeManufacturing]

I posted the same program out, without the circular pattern. then copied that toolpath, and changed the A axis position. worked just fine. so it has to be localized to the circular pattern as far as i can tell.

i did pick the pocket at 180 degrees instead this time though, because that was the one that gave the proper code on the y axis.

[Anonymous]

I recall you used section of model OD as rotation axis for circular pattern and I believe that alone is the reason for discrepancy, should have used longest feature on model like ID or X axis itself once axis have been properly referenced to model in setup.

[CuttingEdgeManufacturing]

im sorry to drag this out, but how can that be an issue using only a segment of a circle to reference? i mean if that is concentric to the bore you would prefer, what could possibly change? is it just fusion has a hard time finding the true center of that circle possibly? 

[Anonymous]

If there is no discrepancy in model design and all features are valid, then theoretically there should be no issue.

However, If I have a choice of using .5" long sketch line or 5." long sketch line to align feature or action against it, I will use long line as that enhances chance for things to be as intended.

Let's say my .5 long line was not properly constrained in course of making a model and it is .0003 off parallel with longest feature on model.

If I use that line in setup, all my programming work can be invalidated, why take a chance ?

[CuttingEdgeManufacturing]

i totally understand. same as indicating in a lathe. always indicate the largest diameter and the largest face or longest feature.

[CuttingEdgeManufacturing]

i tested your theory and made a circular pattern around the center bore. and got the same issue by checking the g code.

[Anonymous]

The theory is that by using more accurate features you avoid potential problems, that seems to not be the case here but both features are tied to how you define setup so I'd be inclined to suspect setup as next potential reason or post not interpreting something correctly.

There is no mystery here, problem lies somewhere in the process, we just haven't discovered it yet.

I have used circular pattern numerous times and never came across anything like it, then again most of that work was done out of bar stock in single setup on mill-turn, eliminating fixtures and potential problems tied to transfer.

[CuttingEdgeManufacturing]

im with ya here. it is a good standard to go by, using the largest possible reference, or at least choosing a reference carefully. 

ill have to dig in more if i get more time.

there was only 3 parts to run so its already done and over with at this point. i appreciate you taking the time to help dive into this.

[Anonymous]

I am temporarily out of work, waiting on stock in mayor project so tinkering on forum is my pass time, I find it better than watching garbage on TV.

[CuttingEdgeManufacturing]

sure is. hones the skills too i would imagine. ive learned a great deal of things surfing the forums here and a few other places over the years.