Uneven surface with 3D model cutting

[Antti]

Hi,


I am using Aspire version 8.024 and ShopBot Buddy PRS CNC. I have an “uneven - not smooth” surface problem, when I am working with 3D model (the original model is smooth). I am not sure is this: Aspire, ShopBot post processor or ShopBot ramp value settings problem?

Anyhow the problem according to my understanding is as follows: originally smooth 3D surface (smooth topographic curve) is converted to small incremental parts of straigth lines. This phenomena combined to ShopBot settings, that the ramp function is not activated , leads to a cutting process resembling more like “jerky large amplitude vibration” rather than smooth curve cutting. I think this is not healthy for the CNC machine itself and does not lead to good surface quality. If the CNC machine ramp function activation limit is lowered so, that ramp function would be activate, I am afraid that cutting process will become very slow and leads to premature tool bit wear and heating? Could You give some advise, what am I doing wrong?



Please find attached an image of the routed surface. I couldn't attach the original Aspire file.
Kindly yours,
Antti

27451

[garyb]

I assume the top of the model is smooth, you cropped that pat of the image.
Its your toolpath stepover.

Stepover’s are measured horizontally in the x and y plane so when you have sharp increase in wall height with the same stepover amount then more cusping from the tool diameter will show up.
Either change your stepover or your toolpath strategy
Gary

[Antti]

Hi Gary,



Thank You for responding. I do understand, what You mean by the smaller step over regarding the tool marks. It is true, that smaller step over with ball nose tool bit in steep vertical surface will make the tool marks smaller, but will it not make the cutting process smoother?

The primary problem is the jerky cutting process within one “topographic curve” on the 3D surface. The surface quality is more like a secondary result. Each topographic curve “feels” like a bunch of small incremental straight lines with sharp corners between – I am not sure if I am able to explain the problem clearly? One can see the horizontal cutting marks (corners) in the image. When standing besides ShopBot during the cutting, one can feel and hear the pounding “all over”. I feel sorry for my ShopBot :)



Any ideas?
Regards,
Antti

[willnewton]

It "sounds" like you are making too deep of a cut in one pass. Or you are stepping over too much. Or you are cutting too fast. Or possibly the spindle speed is too low. Can you tell us the size and style of your cutter? That will help.

As for the machine cutting segments or facets rather than smooth curves. That may be due to your 3d model having a low polygon count. The fewer points there are to sample, the less detailed the resulting model will be. Did you generate the model in Aspire or import it from another source?

[cowboy1296]

pm sent with my email if you want me to look at it.

[Antti]

Hi Will,



I would assume the cut depth is OK, less than 0.1 inch. I also have a feeling, that the primary issue here is the low polygon count as You mentioned - not the step over distance. I made some tests and the step over distance has no effect the vibration or pounding.

The 3D model is imported to Aspire as a STL file. The STL model however is smooth. Please find attached images about the STL. There is one Rhino surface analysis image and one screen capture from Aspire imported 3D model.

My feeling is, that Aspire is creating the roughness. There are too few points on each specific topographic cutting curve (each Z-level), when toolpaths are being calculated. Do You have any idea, how can the number of sample points be increased to achieve higher toolpath resolution?

The tool used is 0.4 inch ball nose bit.



27492 27494

Regards,
Antti

[cowboy1296]

when cut with my tools it cuts smooth. The problem might be in your tool path data base for your finishing tool.

Try calculating a finishing with a smaller ball nose. I primarily use 1/8 and 1/16 and have a 1/4 but have never had any luck with it.

[Burkhardt]

Is this the roughing tool path or the result of the finish cut removing only a thin skin?

Not sure if related, but when when cutting such kinds of surfaces with a large diameter ballnose sometimes I get a lot of chatter, especially when the bit is long and the entire tip of the ballnose is in contact with the material (i.e. roughing or a single rough/finish cut). I suspect the geometry of the round tip makes the bit/spindle flex and "bounce" in the cut groove.

I never have this with an endmill and also not when a ballnose is used for finishing and removing a thin skin only. For a single rough/finish cut it usually works better cutting a perimeter groove first and then from "bottom up".

[Antti]

Hi,


The fact that Your machine is cutting smoothly is in line with ShopBot support opinion. There are some machines, which has this problem. The reason might be slow data exchange speed from computer (controller) to another. I have been analyzing the response times, but so far no luck. Response times have been well within limits <20 ms.

So far the best improvement to this situation has been achieved by changing the ramping values. Could You give me the ramping parameter values of Your ShopBot [VR] command. I would like to compare Your values to mine.



Regards,
Antti

[bleeth]

This could also be a design resolution issue. Too low a resolution setting of the model to start with can make some pretty ugly cutting.

[Antti]

Hi,


This is 10 mm ball nose (100 mm long) finishing cut after 16 mm end mill rough cut with 2 mm allowance between (skin only). So.. only small fraction of the ball is in contact with the surface. I have forced the cutting from top to bottom due to the work piece rigidity demands (tabs at the bottom are cut thinner at the end). I have to try the bottom to top cutting.


P.S. How do calculate single rough/finish toolpath with Aspire 8?

Thanks,
Antti

[srwtlc]

What finish toolpath strategy are you using? If you use 'Offset', you can get some unwanted results on steep edges. If you did use 'Offset", try using 'Raster' instead with a 8 - 10% stepover.

In the 3D view, click on the toolpath (checkmark it) and then zoom in on the questionable areas to see if you can see a lot of short cuts with short raise up moves.

As far as 3D ramping, start here (http://www.shopbotblog.com/2008/03/a-ramping-the-vr-command-and-how-to-tune-your-tool-for-maximum-performance/) and adjust the values from there to find what feels good for your machine.

[Burkhardt]

Hi,


This is 10 mm ball nose (100 mm long) finishing cut after 16 mm end mill rough cut with 2 mm allowance between (skin only). So.. only small fraction of the ball is in contact with the surface. I have forced the cutting from top to bottom due to the work piece rigidity demands (tabs at the bottom are cut thinner at the end). I have to try the bottom to top cutting.


P.S. How do calculate single rough/finish toolpath with Aspire 8?

Thanks,
Antti

Well, then my chattering theory probably does not apply. Looking at your picture, the chatter marks would look more random and rough, anyway.

The single finish/rough cut makes sense if your flute length is longer than the machining depth. Then you just specify that length as the maximum cut depth for the finish cut. This does only work for very small stepover (that you may want to have anyway for good surface accuracy) and has the advantage that most of the material is removed as a side cut with the entire flute, not just the ball tip.

It does require a starter slot or just starting the cut in air outside the material block. Otherwise the bit will plunge full depth into the material with very unpleasant side effects. Not sure how to do this in Aspire, I use Vcarve and Deskproto.

[Antti]

I would like to express my gratitude for everybody who helped me on this project and special thanks to SRWTLC for his valuable tips and Mr. Brady for his excellent article about ramp value settings. I also received great help once again from ShopBot Support (Tom).

And... once more I realize how little I know and how much there is to learn. [VR] is one of the things I have to learn more deeply. Anyhow I am glad to tell, that the vibration problem has been solved - at least regarding this project. The major problem turned out to be too small Z-movement speed (plunge rate). Too big difference between X/Y- and Z-movement speeds probably caused restricted X/Y-movement during 3D cut. This again led to situation, where ramping was not activated, though it should have been.

Here are some images of the final product. Cutting was so smooth, that less than fair amount of sanding was enough for this surface quality.
Thanks, Antti

27502275032750427505