machining 7075 aluminium on DIY MDF CNC

[mihaelbele]

Hello,

I will try to make a guitar neck out of 70x35x100 mm piece of 7075-T6 aluminum.

I have 10k-30k rpm Kress milling motor installed on DIY MDF CNC machine with keling 425 oz-in steppers. Max feed rate that I can achieve is 600mm/min.

Machine is not very rigid in my opinion. I tried to cut 2mm thick aluminium in one pass and got it running in one direction 250mm/min @ 10k rpm with 6mm single flute endmil, spraying WD40 over the endmill every minute or so. I got into 'standard' aluminium chip gumming problems, but the real problem was cutting in the opposite direction, where machine started jerking until I reduced the feed rate. At certain point i had to reduce the feed rate down to 10mm/minute for the back-x and back-y direction. Once x direction changed again I could increase the feed rate to 250mm/min again. That tells me that I (probably) havent destroyed the endmill, but maybe have problem with fastening the milling motor tight enough...

I will have to buy endmills for doing this new project. I was thinking about getting the fly cutter endmill to do rough cutting of straight parts, U shaped or Ball shaped endmill for shaping the oval parts (eg. back of the neck) and ?-flute endmill for cutting.

What kind of endmills would you suggest I should use in a sense of shape/size, number of flutes and the material (HSS/Carbide), and regarding to that - RPM of the milling motor and pass depth ? And are there any considerations about different endmills for roughing / finishing ?

Also, I still haven't figured out how I'm going to do the headstock which is under 7 deg. angle. I was thinking in doing that last, and securing the billet under 7 deg angle, so I can work on headstock like it's straight. Any ideas on that would also be very helpfull.

thanx

[tahustvedt]

I use Vectric Aspire (www.vectric.com), it's a nice CAM software that gives me good control of milling 3D shapes and would make milling the headstock a piece of cake. Your idea of angling the headstock to mill it flat will work, but I find it very difficult to do accurate work when I do things like that.

I also have a DIY CNC router made from aluminium and MDF and use a Kress FME1050. I use very fast feed rates and shallow pass depth when I work with aluminium. This keeps the tool from clogging and as long as the part isn't small I often don't even have to lubricate. With a 6 mm single flute carbide end mill I would try to run 15000-20000 rpm (2 to 3 on the speed control dial on the Kress) with a feed rate of 1700-2200 mm/minute and pass depth of 0,2 mm, maybe slightly shallower if it's 7075-T6. You can mill at a slower feed rate I guess but I would you run at 10000 rpm to reduce the heat and need to lubricate constantly. I prefer three and four flute endmills myself for aluminium and run three to four times faster than I would with single flute. Here's a little video of the machine milling some aluminium at 6000 mm/minute:

YouTube - CNC-milling a little shaft coupling

YouTube - CNC-milling a little shaft coupling

Rastering back and forth doesn't work well with aluminium because that means you're milling conventionally half of the time. Climb milling is the only way to work with aluminium.

WD40 is great, but messy. I use WD40 and then spray red spirit near the end of the job which both lubricates/cools and cleans everything up nicely. Red spirit (and maybe other spirits) also works great on aluminium but is a bit more smelly. You can see me spray red spirit on the part in the video above.

[mihaelbele]

Thank you VERY much!

I thought that I was an idiot trying to make this work. Your information is precious to me. Before I started working, I did some tests, only to find out that finished piece will be imperfect because of the imperfections of the machine itself. So I gave up and started building new machine using steel construction, profile LM rails, and ballscrews. Here's the sketch how should it look like when finished.

I will try to use the old machine to make aluminium parts required for the new machine, using the information you provided. Thanks agan! Cheers!

[tahustvedt]

Ballscrews will help a lot with achieveing higher speed. My steppers are just 1 Nm (142 oz/in) motors, and I use 5 mm and 10 mm pitch ball screws. I'm replacing the 10 mm pitch ball screw with a 5 mm pitch ball screw now as I'm rebuilding my machine to make it longer.

[mihaelbele]

Does the decrease of the ball screw pitch raise accuracy, and vice versa, does the pitch increase raise the max speed of the ballscrew ? I'll soon have ballscrews of 20mm lead, for y-1505/c7, and for z-1502/c5, and I'll need ~1400mm long ballscrew for x axis (i was thinking of 2005/c7 or 2505/c7). Do you have any comments on how should these perform, or suggestions for specs of x-axis ballscrew ? Thanks.

[tahustvedt]

Less ball screw lead means higher resolution per step. A longer lead means the axis can move faster per ball screw rpm, better with strong motors and long screws. Your high torque motors probably can't spin very fast, so you probably need 20 mm lead. I have never tried stronger motors than the ones I have now though.

With 20 mm lead you'll have 0,0125 mm resolution at 8 microsteps. That's the same resolution I run with. I think your super strong steppers will handle 20 mm lead really well.

[mihaelbele]

Hi, here's a latest sketch...


and I have a question regarding the way lead screw nut drives the table that rest on blocks/linear rails.

this is the Y-Z profile view of the table and the lead screw nut fixture in two versions. When lead screw turns it pushes the nut/fixture/table to the left or to the right. Since table rests on rails, forces that press the table downwards can be ignored, and only inertia has to be counteracted.

I have a feeling that option B is better, but I cannot explain why. Can somebody please tell me if one of these systems is better than the other, and explain why.
Thank you.

[JWB_Machining]

I don't know if I follow exactly what you're asking in your last post but I would go with option B because your table will be much more rigid. With option A the table might potentially bend under the weight, thus throwing off your work. I think that's what your asking.

[mihaelbele]

Originally Posted by JWB_Machining I don't know if I follow exactly what you're asking in your last post but I would go with option B because your table will be much more rigid. With option A the table might potentially bend under the weight, thus throwing off your work. I think that's what your asking.

Thank you for your answer. Yes, in option A, force that moves the table creates a moment that wants to spin the table arround the X axis (creates forces at the Y end points of the table, that push one endpoint up, and the other down). And there's also bigger strain at the point where support is connected to the table. But table rests on the blocks/rails that ensure rigidity and will dissalow bending. I'm interested in how will these options handle table intertia, and why.

[mihaelbele]

Temporary MDF machine. My final goal is steel and aluminium, but I was impatient, so I build MDF one. Here's the test drive:

YouTube - Jadranka

YouTube - Jadranka

I'm currently setting paralelles and 90deg angles bit by bit 'till I get something I can make aluminium parts for new machine
max speed x:9.5m/min, y:8.5m/min, z:12m/min
steppers are 382oz/in, ballscrews are c5 (x,z) and c7(y) 20mm lead.

Check your motors if ordering from kelling inc. I didn't, so one year later, when I dismantled old machine, I had an unpleasent surprise. Instead of 425 oz/in i got 382 oz/in motors a year ago. Figured it out when shaft diameter didn't match the specifications. Wrote to kelling, not a word from them. Doesn't matter, another oportunity to learn.
Check the stuff you buy from ebay. I ordered 8 same hiwin blocks and inspected only one, 6 months later, it turned that I got 4 blocks of one kind, 3 of other, and 1 of third kind. However, they were all the same height, so it's fine (though, one is much noiser that the other seven )
Also one more warning, beware of ebay rails/blocks that are not NIB, I got some that were clear when I got them, and few months later they were becoming rusty - NIB ones didn't.

One big Thank You goes to tahustvedt!