Strategies to Reduce Machine Time

[gene35146]

I just did my first 3D piece (and learned a lot).  It took nearly 19 hours to do all of the machining.  I didn't have any large (1/2") ball nose or straight bits (have ordered them now) and when I put that in it certainly reduce the machine times by a lot.  

I understand the purpose of a 3D roughing pass is to hog out a bunch of material while the finish pass makes it look nice.  Is there a strategy you can use to use multiple bits to do large areas more quickly?

Take for instance the piece I have below. Could I have used the 1.5 mortising bit to hog out the large surface areas first.  I am sure I could have created multiple 3D roughing passes but what is not clear is how they would work together.  For instance if I just picked a big open section and machined it down to close to finished height that would go very fast but when I do another pass with a different bit to get in areas that the larger bit could not get, how would that pass know not to just go over the same area again?  

It is all just about creating multiple vectors and many tool paths?

Also there is a setting on a number of the tool path that essentially says start lower than the top of the surface but I am not sure I understand it as when I use it, I always go lower at the end of the pass than I think I will?

Attachments:

[redwood]

I'm not sure how big that sign is, but 19 hours is a lot. The 3D roughing pass is solely used for protecting the smaller finishing bits, so they don't have to work so hard. So adding multiple roughing passes will actually probably take longer.

With the setting to start your toolpath lower, you have to remember that the second setting is the depth you want it to cut from that starting point, not the overall depth.

[gene35146]

Understanding that 2nd setting is a big help. I actually cut this twice one time with a 3/8 end mill doing the roughing pass and a 3/8 ball nose doing the finish pass. The first pass with the end mill did not turn out as well as the 2nd pass I did with using the 3/8 ball nose for the roughing and the finish pass. I know one of the reasons the 19 hours is it takes forever with a ball nose bit to hog out large areas. I also suspect that getting the perfect height on the ball nose on the 2nd path helped as you have to set the bit height again if you use different bits. I will just keep on learning I guess.

[traindriver]

My longest machining project took about 3 hours to rough with a 1/4" end mill, and 20 hours to finish with a 1/8" ball nose and a 6% stepover (set in the tool database under the "Cutting Parameters" heading in case you didn't know) and 50 in/min feedrate. It is 18" x 18" out of 8/4 oak. The first picture is how it came off the machine (except I cut the octagon on the tablesaw) and the second is how my friend chose to stain it.  Oak is pretty grainy, so it doesn't show tool marks that well.  The third picture is one of the first things I did, and if you look closely, you'll see the tool marks from where the machine made widening circles cutting out the turkey.  It is especially noticeable around the edges of the carving.  I went to the woodworking expo in Atlanta last year and talked with some guys from Vectric and asked how they got their displays so smooth - did they come out of the machine that way, or did they sand them a lot.  He said that they came out of the machine that way.  He then asked how long it took to machine, and how long it took to sand.  About an hour machining, and 30 minutes or so trying to remove tool marks was my answer, and he said that my stepover was likely too large, which is why I had the big tool marks.  SO, you can make a small test part and look at your finish, then play with the stepover to gain a little bit of speed (adding 1% to the stepover reduces the finish cut time from 20 hrs to 17 according to VCarve Pro), but keep an eye out for the tool marks.  You can also play with the feed rate.  I don't know much about speeds and feeds, so I leave mine at the default and it does *okay* but sometimes I burn the bit a little, which a machinist friend tells me means I should increase the feed rate.  I have seen on some posts here references to speed and feed info on the internet, so you might try googling.

[traindriver]

I made this sign that required both v-carving and 3d carving.  I didn't want to waste time machining the area of the whole sign with my tiny 1/8" ball nose, so instead, I did a roughing pass with a 1/4" end mill, then when I made the finish toolpath, I set the Machining Limit Boundary to be the Model Boundary, rather than the Material Boundary.  This DRASTICALLY reduced the machining time, because the ball nose only covered the part I needed it to, since everything else was already flat from the roughing cut.  The only thing about doing this is that there is a small outline around the model, since the roughing cut leaves a little bit of material for the finish cut to take off.  An additional advantage, though, is that you don't have to worry about small tool marks from the ball nose in the large areas that are just flat.  Since the sign was painted, the small trough around the border of the model doesn't show up.





One final alternative would be instead of having the model raised, you can have the model "sunk into" the wood.  I was playing around one evening and came up with something that I posted in the link below.  Since the model is below the surface of the wood, your rough and finish cuts don't have to cover the entire area of the material, and are much faster.  Also, it's kind of a unique effect in that the face of your bear will seem to be looking at you from any angle.  Hope this has given you some ideas to try.

axiomprecision.proboards.com/thread/172/inverted-3d-model

[Deleted]

19 hours sounds long and there are likely things you can do to cut that dramatically. It looks like your design is almost all 3D, including the text, that does slow things down but it also doesn't look like it's all that deep so you could likely just do a finish step for the whole sign. The purpose of the roughing is to remove the bulk of the material so that a finish step is not removing too much and can go at a high speed because it has a finer step over.

There are some tricks you can do in the roughing stage that makes a big difference. Don't remove extra material, meaning don't go to the edge of the board. Take your outline and offset it by a little over your rough tool width and use that line to set your machine to Selected Vector, it may smooth out the edge of some detailed areas, and you can always dial in some boundary offset. You may be able to just use the model boundary with an offset but that depends on your project.

To save a tool change you can use a ball mill for roughing and finish but it may make sense to tell the software the roughing pass has an end mill because the step over will be larger because the software is assuming a flat bottom and it doesn't want to leave large scallops. I've seen this make a 3x time difference but if you're machining flat areas actually using the end mill is desirable if the finish pass is not going to cover that area. If the finish pass will be covering that area then the large groves the ball mill will leave don't matter. At the same time, you can edit the tool step over for roughing passes, it may give you some warning messages but if you think you are being reasonable they sometimes can be ignored.

Using Z Level strategy without a profile step can be faster than 3D Raster.

Choose the appropriate ball or taper end mill size for your detail. Going too small may increase the detail marginally but the machine time by a large amount. Render it with several size tools and see if it really matters to you. Turn up the Preview Simulation Quality before you do so that you get a better comparison. You can even Save Preview Image and then display them side by side on your screen and see if it matters. For this step, you may notice the difference so asking someone else which they prefer or if they can tell the difference may give you a more objective opinion.

Sometimes it makes sense to add extra vectors to define areas to use as Select Vector limits. In your project, you could use this technique and a pocket path to do your large flat area quickly and then do the text, bear and border as three separate 3D areas, perhaps done at the same time by outputting multiple tool paths to common files for rough and finish. Using this technique and just a finish pass for the 3D areas, I would think you could get your total time down substantially.

Generally, if your times are high, play around with some other strategies as you can easily get many multiple speed improvements.