Machining off center

[Anonymous]

So yes, this topic has been covered , but no solutions that I can see.

I have been busy trying to make this part for quite some time and have got a handle on most of it.

So now have machined some soft jaws for drilling the holes in the sides and flip over to machine from top down.

The issue is not the machine, I have been doing these individual operations for some time.

I can drill and counter bore the holes and repeat it every time. The second op is where it goes wrong, but never used to.

The odd thing here, is the center of the case (watch case) is still center, but the tool path seems to offset the rest by a few mm. I have ruined $100 worth of stock so far in the last couple of days and am left confused.

I have tried 2d and 3d adapt clear on this and same result. I verify the WCS is correct and the first OP's in G54 are fine and where I expect them to be.

This second OP in G55 is also according to the machine,and only the other day was working fine. I tweaked the tool path, but do not remember what I did and it is not working now and destroys the part.

Basically, as you look at the part, it machines with a bias to the left.

The machine is German made and is running Mach3 ,in very good condition with ball screws and linear guides and repeats on other tool paths just fine. Table is square and so is the fixture and stock. Haimer 3D taster confirms all this.

I have been working on this (new to CNC) for a long time and almost have made that final step and now this!

I post the part here , and if anyone can take a look and understand where I am going wrong, that would be much appreciated.  Other tool paths seem fine, it is just the adapt clear , though I have not yet used a different path such as scallop , to remove all the stock to see if that is different. That is my next move after asking here.  

[HughesTooling]

How are you centring the body in the stock and which way are you flipping it over. Is it the 2 ears on one side, the part not being symmetrical that's causing the problem? Why not create a component in the model workspace with your stock so you know your reference will be the same when flipped over. 

You only have one setup in the model attached so can't tell how you set the other side.

Mark

[Anonymous]

I ordered the stock , rather than machine it flat and square. 

I made soft jaws and for the first operation , machined the jaws to hold the stock on its side dead center in the jaws and then also created an area around that, to flip that over flat looking down from the top. It is all center in the jaws and both where machined at the same time. I have already machined a successful piece a week ago or so in that fixture.

G54 is fine and repeatable. G55 went wrong on the first try, (same issue as now) then I put in a piece of stock without doing the G54 operations and it was fine.  Have it in front of me now. Then I tried again and it was wrong.
So I re-did the tool path just in case and wrong and today I re-did it again. It appears in the simulation to be ok but wrong on the machine.

The setup in the model attached is a stand alone one to try and fix the issue away from the original lot, which has not worked. 

Basically, I stand it up and do the side that has three holes and spin it around to do the other two holes. All are where they are supposed to be.  I lay it over (re-confirmed the WCS today) and start to machine from top down and it goes wrong, but the center is ok, as it will actually put the hole in the middle where it is supposed to be. When the tool path gets down to approx (not sure exactly) -10mm or so, it comes correct!

[HughesTooling]

I find it hard to know what you are doing when you flip it over and look down from the top doesn't mean much as I don't know what you're calling top. Makes sense to you you've been working on it for weeks but means nothing if I'm not looking over your shoulder.

Need the file with all setups really. Comes down to origins not matching in Fusion or your machine or both or could be a bug but I'd be looking at the origins and how the body is centred in the stock. If you draw the stock in the model workspace and select the same body in all setups and select the correct corner at least you know 100% the stock and model stay in the same relationship.

Mark

[Anonymous]

I tried to export the file, but it wants to do only IGS etc, not the format 3fd or what it was before.

Below is the full setup , though not in a file you can read of course. Left is G54 (the holes ) middle is G56 and right is the ''top down '' I mentioned , which is G55. 

The file I posted in the OP, is the issue , as I can do G54 , skip G55 and do G56 instead and it works. 

WCS is all correct and the machine repeats as I mentioned above. My thinking is, it must be in the post, but I can not understand why

[HughesTooling]

When you run the file from the first post on it's own does it still have the problem? Setting up a drill op just as a test it shows the offset from the stock origin to the hole centre is X29.25 Y-27.5, is that correct and do you get the same from your post?

Mark

[HughesTooling]

As a test I set the tolerance in your Adaptive cut to 0.005 and fitting to 0.05 increased the stepdown to 3mm so the calculation's a bit quicker and posted out using my post for an Anilam control and get this, circle centre matches above. It was still pretty close with your settings just a lot more code.

Mark

[Anonymous]

Yes I do see X29.25 Y-27.5 if I set up to drill that one in the center.

The first post however is on its side and has no material to drill a hole dead center to check, but it would only make the holes misaligned relative to the adapt clear of the part if something was wrong there.

What I am seeing if I run the second op from a view that you posted , is that X29.25 Y-27.5 center of the part and it then machines with a bias to the left. 

The other day I threw another piece of stock in and it machined it ok. Very strange, but now I am 3 in a row for machining left and need to understand why before I destroy more stock. I can do other parts without issue, so I can safely assume it is not the machine itself

[Anonymous]

So of course it should be X27.5 and Y27.5 , as the block is 55mm by  55mm .
I just modeled the stock up as a solid and found the center point to drill a test hole, while putting the stock in exactly the same spot as the stock for the watch case. Same stock after all and same fixture

I get a X27.5 and a Y27.5 in the code, on the machine also and looks center by eye

I use the 3D taster to confirm my WCS is still correct and it is.

So I am thinking it must be in the tool path that I listed and that uses the same stock and WCS , G55 .

Given it is the third tool path and I think 4th failure from 5 attempts, something has happened to my Adapt clearing as far as I can see. 

I have a little too much stock to clear without it, as these are Stainless parts (though I am currently practising in Alu)

Can I ask also, what program is that you have used to show the center? That might be helpful for the future 

[Anonymous]

ok, so think I found it.

In the setup it all says its center , but it is not.

It is a fixed size box 55mmx55mmx25mm and the model has everything selected as ''body'' 

The offsets are all center, except the height , which I need different. 

I can see the center point is off to the left there, so that will explain it, but how to fix this? It was ok previously 

[HughesTooling · Accepted solution]

Like i said in my first reply the 2 ears on one side make it non symmetrical. If you draw the stock yourself in the model workspace you can make sure it's always correct.

Mark

[Anonymous]

Sorry, I didn't get that. I do now however.

While I have you, if I draw the stock, how do I then see my part?  I know it is obvious, but for some reason I don't quite see how I am going to make that work with my part to put tool paths in . 

I am sketching a rectangle with a center point on the bottom plane and not extruding anything, that is all I can think of to try.

And thanks by the way

[HughesTooling]

I usually make a component called stock so it's easy to hide\unhide. When you create the setup have the stock hidden so only the part you're machining is selected then on the stock tab select from solid for the Mode unhide the stock component and select.

Mark

[Anonymous]

OK thanks.

The way I did it seems to be working, as it is in the machine at the moment and looks ok as a simulation. 

But I will do it the way you do on the final setups. 

Cheers for the help

[HughesTooling]

Don't know if you're interested in reducing the code size but setting the tolerance to 0.005mm and smoothing to 0.02mm the code is reduced from 1.9m to 1.0m. The toolpath will be smoother as well. The disadvantage is it will take longer to calculate.

Mark

[Anonymous]

Absolutely interested, thanks! 

I would have thought it would have had a lot more code, but I will do it now. Don't mind some extra caculation time