Surface Blend Stepover issue

[specv]

Hello all,

I am cutting a impeller floor using the blend toolpath, the issue i am having is the stepover is not consistent. as you can see from the screenshot its obvious. The settings are all the same, the tool path was copied down and given the same surfaces and check surfaces, but different chains for blend. Any Ideas??

 

[djstedman]

Not sure what would cause it, I have run into similar things before where certain geometry just didn't want to cooperate.

 

If the features are in fact identical, then how about perhaps doing a transform rotate, or making copies of the model that are rotated.. probably easier than fighting with Mastercam not liking something about that geometry.

[CJep]

When programing impeller or compressor type geom I program one section then rotate the toolpath.

If you use the advanced multiaxis toolpaths the rotate function is in the roughing utilities page.

[Tinyfxds]

Could there be something wrong with the surfaces on your model?  I've seen some weird things happen when surfaces gets twisted.   

[specv]

I can't see it being the surfaces. The stepover is the stepover regardless of what the surface does.

[MotorCityMinion]

I can't see it being the surfaces. The stepover is the stepover regardless of what the surface does.

The blend tool path will compress the actual tool path to cram an equal number of passes in between the blend chains, not necessarily using a even step over. If the surface boundaries change, your blend chains should follow suit. Variations in the surface will have an effect.

 

The drastic changes in the tool path pictured is odd. Surface normal perhaps?

[specv]

its 1 surface, I understand the toolpath shrinks the stepover, but its programmed to .009 and there are areas where its .035. 

[CJep]

The change in stepover looks to be relative to where the cut is running in the geom, I believe this has more to do with how the blend is reading the associated geom. The blend toolpath is intended to morph between two chains which is what is happening at 4:00 & 10:00. As the toolpath transitions to the other segments the blend is starting to morph in more of a radial fashion.

What are you using for the blend curves?

 

Also as stated earlier I'm certain you will get better results programming for one of the segments then transform rotating the toolpath to cut the rest of the part.

[specv]

There are 8 segments, 2 sets of 4. The curve in each vane is my blend curve, this is one toolpath per section. so the surface and chains that are used are identical, just copied/rotated.

[specv]

The 10 and 4 oclock segments are the ones in question, the largest stepover on those 2 segments is .035 but programmed to .009. The ones at about noon and six are right on the .009 stepover. I had to change my stepover to .003 to yield a .009 on the 10 and 4 oclock segments.

[CJep]

Do you have access to the multiaxis toolpaths?

I think morph between two curves will give you better results?

That toolpath is blend with a crapton of control.

[specv]

I do have multi axis toolpaths, but was told by my reseller not to use it in 3 axis applications. Blend between curves should work the same for 3 axis, i would think. This seems like a software issue to me. I'll try to send something to mastercam.

[MIL-TFP-41]

Can you share the file? Or at least just the portion in question?

[specv]

I cannot share the file, I did try one test though. I program for a horizontal so this was using the front plane. I had an idea to take the part flip it to top plane and regen, and BAM! perfect stepover, no issues whatsoever. looks to be a bug to me.

[CJep]

I do have multi axis toolpaths, but was told by my reseller not to use it in 3 axis applications. Blend between curves should work the same for 3 axis, i would think. This seems like a software issue to me. I'll try to send something to mastercam.

 

Sure you can lock the multiaxis toolpaths to 3 axis.

Here is a morph between curves, rotated in the toolpath, locked to 3 axis.

I think the issue you were seeing in the blend had more to do with an algorithm based on toolplane, I've seen strange things happen trying to use a flowline tpath in a rotated plane as well.

[specv]

I sent it to CNC software they logged it as a bug. 

[K2csq7]

For anyone that would like to keep an eye on this, we'll be posting any updates here....

http://forum.mastercam.com/Topic3929.aspx