LR4 Aluminum Milling

I’m testing parameters, no calculation this time, only experimental testing.
I started from @DougJoseph Feeds&Speeds seen in one of his videos, an worked on them, I think I’m going in the right direction:


This plate is 6060 5mm thickness done in a single pass with trochoidal milling path.

Here my parameters:

RPM are wrong, I’m using 12000 (Makita dial 2)

I tested both Conventional and Climb milling, and I found that Climb seems to provide a better finish
I took some pictures, and the difference is not that much, but with bare eyes it’s more noticeable

CLIMB:

CONVENTIONAL

Conventional on top

6 Likes
7 Likes

Wow! These cuts look very good!

2 Likes

You mean the ones that he took from @Jonathjon who took inspiration from @azab2c’s collection and my early tests? :stuck_out_tongue:

5 Likes

“Stolen Values” coming soon :joy:

3 Likes

To be honest I only saw @DougJoseph ones in the video were he was milling the LR4’s XZ Plates

3 Likes

Looking good! Yes, for what I tried I indeed looked to others who had already done successful cutting in aluminum. I did make some tweaks, but at this time, I don’t remember what they were.

1 Like

That matches what I remember from machine shop class long ago.

How accurate are the inside dimensions of those slots vs nominal?
I haven’t found a great source on belt stiffnesses, but since we’re using fiberglass filled they will be pretty stretchy when cutting AL vs wood

Climb prevent the rubbing of the tool at the beginning of each cut, but it’s more demanding for the machine rigidity, so in our case it can be done only with not too aggressive cuts to prevent chatter.

Nominal 18.5mm

Yes that’s my concerning, A lil more tension can help, but if too much tend to create problems.
To prevent it I’m building a LR4 with 400x400 bed, to reduce deformations and the belts’ length.

1 Like

@Rico_LRS
Under size is what I’d have expected.
When the tool was cutting each side it was pushed slightly away by the metal and so didn’t cut as much as possible

More tension would probably only help if there was slack in your belt while it was cutting.
These belts are going to behave linearly. So their spring rate won’t change as a function of tension.
The only significant way to change stiffness is to use shorter belts or belts with steel core but the manual specifically says not to use those.

2 Likes

You can certainly try but you’ll see pretty quickly why it’s a bad idea. :stuck_out_tongue:

The flex can be mitigated by going more slowly or using a finishing pass.

2 Likes

I’m thinking about Aramid core belts…
anyway this big LR4 is intended for wood and soft material, metal will be milled with the “stubby” one.

2 Likes

I already use it, and shallow one… maybe with a better CAM you can add things like tool compensation, or backlash compensation to deal with it

1 Like

Ooh that’d be lovely - imagine a material tool test - cut a square of designated size, measure actual size - enter into the software and it accounts for the discrepancy…

One can dream…

1 Like

And it’s a amazing machine which is punching way above its weight class.

Any software compensation would be very hard to implement, estimated cutting forces, upcut vs downcut, which side of the tool is cutting, axis position would all play into it.

1 Like

Industrial grade cncs have correction parameters.
Anyway, it’s simple than you think, you can reduce the bit diameter into cam setting, that’s it

1 Like

The finishing pass is more or less always the same for a given tool, You can test it and modify the tool diameter into cam settings

Have you tested both ways? You might be surprised where the machine stands in rigidity.

Did you do a full depth finishing pass, is so what amount of material did you leave.

Really good idea not to mention those, they 100% do not work under any circumstance for our machines. They fail extremely fast minutes to hours at most. If people read they are better in anyway they will want to try them.

Fairly certain they are not fiberglass cores any longer. Have you checked?

It might be too shallow, the tool does need to bite in metal.

Some people use two finishing passes instead of one.

You guys are talking in certainties here and I am not sure I agree with all of these statements.

2 Likes

Just remember these are not 2 ton cast iron machines, they do not handle like each other. Some things are universal truths and some are not. With a little testing, you should be able to get extremely accurate dimensions.

Your surface finish is amazing and if all your cuts are off by the same amount your issues can easily be fixed with different CAM settings.

Poor quality comes from lack of precision, not accuracy. To me your precision looks amazing, from there you can dial in the accuracy.

2 Likes

Yes, and I said that climb give a slight better result

Climb seem to solve this problem, but the topic was intended to report progresses about aluminum milling, more testing will follow.

I’m very happy about the results, I’m sure I’ll solve the dimensional inaccuracy, it’s not a big issue, we had them also with 150k€ Mazak 5 axis machine

3 Likes