toolPath to cut 2 pockets layer by layer

[CuttingEdgeManufacturing]

hey guys,

got a part that came back to haunt me.. long story short we have 60 parts to do, that are made of PVC plastic. i think about 95% of the volume of the stock is removed. leaving some pretty weak points in the part. last time we tried to make these parts we had the long skinny webs break off when finishing the pockets...

basically my strategy with these was to leave the part as solid as i could until the last bit. so i pretty much get the part complete before i put the pockets in, then do the tiny tabs at the end of the part.

anyways, does anyone have tips to machine these pockets? i figured ifi  just step the pockets down .025" at a time doing 1 pocket down .025 then going to the other pocket, effectively finishing the web while leaving it as rigid as possible. but i havent found a good way just yet.

looking for overall tips on this one.

[Anonymous]

First thought that hit me was "order by depth", without having a clue as to what your model looks like, here is one raw example where tool alternates between two pockets to cut one level at a time.

[CuttingEdgeManufacturing]

what toolpath is that under? ive tried 2d pocket, and did whatever setting that had for order by depth. but it didnt really work as expected?

sorry i forgot to attach a screenshoot yesterday. here is the part in question

[Anonymous]

Look at the sample file I uploaded earlier, 2D pocket with "order by depth" selected in cutting tab.

But, looking at screenshot of your model, this may not be the option because pockets are not on same plane or if it works, you will have a lot of A axis spinning back and forth for every pass in each window, just an observation at the glance.

[CuttingEdgeManufacturing]

sorry should have mentioned i got it figured out. it appears as long as your selections are all on the same plane it will work as intended. if i selected the bottom of these pockets, which put them on different planes, then it doesnt work right.

also no 4th axis. currently this is 5 ops in a 3 axis machine.

[Anonymous]

You could make a sketches that divide windows in half, start cutting each pocket in the middle and drive tool towards the side, making passes parallel to thick side plate. 

Using 2D Contour with passes and depths, you would have stock in half the pocked provide rigidity while load on  thin membranes is diminished as tool approaches side, also benefiting from stock on depths up until last pass, do that on both windows and then do the other half on both windows, starting from middle and driving tool towards the side.

[Anonymous]

Just thinking out loud here, why not do the windows while below stock is intact, then remove that stock while holding part in a way that doesn't stress the weak points?

[CuttingEdgeManufacturing]

not a terrible idea, my main concern with that approach is the fact that the stock will be rigid at the start of the program but at the end it will be pretty weak so it will likely buckle during operation. 

the way im planning currently we hog out alot of material while its a rigid piece of material, then go in and slowly remove the windows. just seems more efficient to me, but hey ive been wrong on this project already. we did these parts like 2 years ago, only ever shipped half of them and kinda just gave up. it was a disaster of a job. absolute disaster..

but i only had about 8 months of programming at that time, and maybe a year in the field of machining entirely. i was in way over my head at that time. 

[Anonymous]

We all have our nightmare jobs to deal with and here, without being familiar with all the specifics, you can only hope to get a hint and possible lead to better option for anything that is not directly tied to how we use Fusion, one common denominator in sea of endless possibilities, sometimes I wish I had other people's "problems" and problems.

My shop is aiming to be the only US supplier of ultra precision part for Hitachi, so far we beat the Japan in level of success but tolerances are so far out of mind that all parts are out of print but they function just fine, it drives me insane.

In the latest twist, one part was scrapped in Mexico, doing roughing operation, as it stands, mistakes in first operation carry over to last operation with gradually diminished margin of error that cannot be corrected if deviating from average by large scale.

Now looking to eliminate outside influence by having stock prepped to solid donut with plenty of stock left for in house processes.

It's another day in paradise, and the show must go on. 😁