enter at midpoint, no consistency in an array

[Dana L Johnson]

I am cutting small rectangular windows in .062 brass sheet with typically 200-300 windows per sheet. the windows are long and narrow and I cut them out using a .032 end mill treating them as a countour rather than a slot to reduce the metal machined. as a result there is a narrow slug in the center of the window that falls away at the end of the cut. I need to have the cutter enter at the midpoint of the long side in it's lead in arc / ramp but I haven't foound a way to force MasterCam X6 to choose the long side. No sort option I've tried treats all the windows the same, infact some tool paths are pretty squirrely leaving out random windows and coming back to them later (or not at all ). For individual windows, I can choose the entry point, but for sheets of hundreds of these small things, it is incredibly time consuming and prone to error. Using window selection is the fast way, but then Mastercam chooses which side to enter at the midpoint, and even in a totally regular aray, ( and many are made of different sizes of windows ) there is no consistent uniformity in it's choice. occasionally a sort pattern will yeild good results, but more often, entry points will be some odd mix of short and long sides. It wouldn't matter except short sides mean cutter breakage sometimes 3-4 per sheet. At $20 a pop, that adds up really fast. Any ideas, my reseller tech guy is looking for a solution, but I thought I'd try here. Wasn't obvious to him and he teaches the thing. I use MasterCam X6 Mill level 2

[Guyinthedesert]

can you try using an entry point?

[brandon b]

In the lead in page is " start at midpoint" selected?

 

Also in the operations manager click on "geometry" then right click on each chain, then click start point and you can control each chain individually. I like to use the dynamic option . It's a little time consuming but hey it's a solution.

 

Maybe first try selecting chains one at a time instead of window select. Make sure you select them all going in the same direction

[Dana L Johnson]

Start at midpoint is selected, but it chooses which side's midpoint to start at. I can try the first suggestion which sounds like at least it doesn't require re chaining, but selecting the chains one at a time is what I used to do and with hundreds of small chains, it can take hours to do. We tried working with entry points, but the tool moves in a straight line from there to the start of the chain. I'll read up on dynamic selection, but I was trying to get away from the one at a time thing. Fortunately window selection handles the accompanying array of tiny holes just fine. As well as the outer perimeter contour which doesn't need to start at a midpoint.

Thanks for your thoughts

[kostiuk]

Just a thought, are you absolutely positive the geometry is all the same? Maybe try replacing one of the rectangles that is starting on the end with one that starts on the side and see if it makes any difference.

 

Kevin K.

[brandon b]

Can you upload a sample file?

[brandon b]

I think I got it. In the chaining popup window "!" Button and activate start chain at point. Give me a minute and ill upload a picture

[brandon b]

this is basically what your trying to accomplish?

[CJep]

You could turn off the start at mid point in lead in/out then break the geom at the mid point you would like to have the tpath start from.

Instead of window selecting the chains have you tried using feature chain?

[Dana L Johnson]

Thanks all for your ideas. The start chain at point may be helpful if I don't have to choose a point for each chain. I need the tangent entry arc and the tech guy at Prism tried breaking the contour but it didn't get us anywhere. The uploaded thumbnail is essentially what I need to do, but the start point needs always to be at the midpoint of the long side of the rectangle. For reference, my array is made of rows of rectangles each roughly 1 inch in the Y direction by .125" in the X direction. The smallest is more like .7" by .062" . They are arranged in horizontal rows soldier style, of from 16- 22 per row. Each row has only one size rectangle. I finally had success making a separate operation for each row, but only about half the rows would start at the mid point of the long side. Even if I selected that as the start point. The problem rows would enter at the midpoint and immediately go to the short side of all the rest of the row. At least half were uncooperative, even rows that were identical except for starting a row up in the Y direction. No rhyme or reason I could figure out. I did discover that for the bad rows, if I chose my start point on the short side of the correct end of the row, all the entry points would jump to the middle of the long side . Once I found the correct end of each row ( it varied with no apparent pattern) I could get the toolpath where I wanted it. Fortunately the parameter settings carry over from the previous operation if it was the same type, so doing each row separately didn't take long. There is clearly a lot of power for recognizing similar objects ( not if they are non standard shapes) but some of it's tool paths leave you scratching your head. I' rather not have to resort to tricks. Thanks again everyone, I think I have a workable if not ideal solution now.

[brandon b]

After playing around with this. I'm starting to feel your frustration, trying to get it to start on the long side.

[brandon b]

After playing around with this. I'm starting to feel your frustration, trying to get it to start on the long side.

[brandon b]

If each row is the exact same shape/size. You can use transform toolpath. Just chain the first one how you want and the translate that toolpath over. I got it to work.

[brandon b]

transform toolpath

[Colin Gilchrist]

Transform Toolpath would probably give you the best results. As far as I know, the chaining algorithm is looking for entites that are connected at the endpoints. You can force a chaining direction (CW / CCW), but the algorithm doesn't have the ability to determine a "long" or "short" side. It just looks for wireframe entities, and then tries to chain them if the endpoints are connected. Once the chain ends up back at the start point, it knows the chain is complete, and begins searching for the next shape to chain.

 

If the spacing is the same between each cutout, you can chain the first rectangle with a Contour Op, then Transform that Toolpath in X or Y to get either a row or column of the windows done. Then you can use a 2nd Transform Op, and translate the 1st Transform Op in the other direction. You could be done in about 5 minutes once you are familiar with how the Transform Ops work.

 

Hope that helps,

 

Colin

[Colin Gilchrist]

Transform Toolpath would probably give you the best results. As far as I know, the chaining algorithm is looking for entites that are connected at the endpoints. You can force a chaining direction (CW / CCW), but the algorithm doesn't have the ability to determine a "long" or "short" side. It just looks for wireframe entities, and then tries to chain them if the endpoints are connected. Once the chain ends up back at the start point, it knows the chain is complete, and begins searching for the next shape to chain.

 

If the spacing is the same between each cutout, you can chain the first rectangle with a Contour Op, then Transform that Toolpath in X or Y to get either a row or column of the windows done. Then you can use a 2nd Transform Op, and translate the 1st Transform Op in the other direction. You could be done in about 5 minutes once you are familiar with how the Transform Ops work.

 

Hope that helps,

 

Colin