My adventures with a bench top CNC mill

[JasonBModerator @jasonb]

A bit more about finish cuts which may be of use to you John in the future when deciding what to use.

Like Julie I often use what F360 calls an adaptive cut to rough out most of the waste. As the tool is making a series of cuts back & forth you wo end up with scallops or cusps when doing internal spaces or fillets when cutting away the material outside of a feature. In this case a follow up contour is generally needed as that will in the case of the two oblong cutouts on the clamps follow a half circle at the ends and a straight line between.

I did not try an adaptive for these two features but based on trying it for the sloping areas where it was taking 3-4 times as long as a pocket due to the small cuts this machine is going to be limited too and the fact that running the free version of F360 means no rapid moves which you tend to get more of with adaptive paths I feel the contour and pocket options are likely to be the best bet for this jon on your machine.

My usual method for adaptives which seems to work for my machine is to leave 0.3mm radial material. I then run a contour with one roughing pass at 0.2mm and a finish one that takes the last 0.1mm off. Again given the lack of rigidity of your machine I would think an additional spring pass at the same final setting would be needed to take out any flex in the tool and machine.

But as you say it’s a practice piece and the other thread would have been very long if all options had been covered, 3000 words and 50 images was enough for me. As I said several times without knowing what sort of cuts the machine will be capable off it is impossible to give an exact approach as there are so many variables involved, so what may work on one may just stall another, the helical entry being just one example.

[John HinkleyParticipant @johnhinkley26699]

Jason,

Once again, thanks for the input.  I’ll file those thoughts away for the moment as my brain is beginning to hurt.  I’ve taken on board what Engine Builder has said, as an experienced Carvera owner and this morning I reworked all the operations with a 3.175mm (⅛ inch) bit as supplied by Makera and recommended by Engine Builder.  This will entail replacing the 4mm collet with the original ⅛ inch one, which is a simple job, and means that the 2mm ball end mill for smoothing the ramps will fit straight in as well, just needing the tool height to be verified.

John

 

[Julie AnnParticipant @julieann]

A few more thoughts:

• The shank of the cutter can always be shortened to decrease stickout
• Roughing/finishing doesn’t take more time, the same amount of material is being removed, but generally a better finish will result. While I don’t build to exhibition standards I like things to look right.
• For long features like the slot I wouldn’t ramp down using a helix; I would use a straight ramp. The performance of helical ramps is very dependent upon the ratio of tool diameter to hole diameter. Ideally the tool needs to be about half the hole diameter or less.

Julie

[John HinkleyParticipant @johnhinkley26699]

Good point about shortening the bit, Julie.  Obvious if you think about it, which I didn’t!

When I make something worth looking at, I’ll use the roughing/finishing technique to get a better finish.  Like you, I appreciate the aesthetics of a product in addition to its utility. All good information to know.

John

 

[JasonBModerator @jasonb]

Any more swarf been produced John?

[John HinkleyParticipant @johnhinkley26699]

I decided to take a step back and reworked the tool paths, taking on board the various suggestions that have been made.  As I’m also making videos of my efforts, I need to take that into account as well as it slows me down a bit trying to get decent shots without long, boring repetitive sections. Then, two days ago my desktop computer took it upon itself to fail.  This coincided with a Microsoft update and an update to my video editing suite.  That has taken me two days to sort out but I’m back up and running again now.  I have some footage of the next couple of operations and hope to publish these today or tomorrow.  I’ll put a link in this thread when I do.

Suffice to say, the operations to mill the slots and roughing the slopes have been done and today’s efforts will see the smoothing operation completed.  You won’t be surprised to hear that there were a few stumbles along the way, but that’s all part of my learning experience.  Trouble is, any lesson I learn is usually forgotten in a couple of days.  It took me nearly an hour just to get the computer to connect to the mill by wifi the other day.  When that sort of thing happens, I get frustrated with it and walk away for a while.

Add to that, a number of medical appointments and the delays all build up.

Rest assured, I’m on the case. It may take me a while, but I’ll get there in the end.

John

 

[Julie AnnParticipant @julieann]

It can be frustrating, but hang on in there. By way of encouragement I managed to CNC mill this liquid cooled power electronics heatsink on my CNC mill a few months after I’d bought the mill:

Granted I was experimenting with the CNC mill every day making simple parts. But before starting I had zero experience of CNC milling and zero experience of CAM progamming, albeit I was conversant with 3D CAD.

It just takes perseverance!

Julie

[John HinkleyParticipant @johnhinkley26699]

Thanks for the encouragement and well done, Julie. Impressive stuff. You have not only engineering training and experience on your side, but youth as well. I’ll be happy if I can knock out a few good looking bits and bobs without too much grief.

John

 

[John HinkleyParticipant @johnhinkley26699]

As promised, and despite continuing problems with my PC, I’ve published the latest video in the saga.

No spoilers, you’ll just have to sit through it like I did to find out what actually happened. (Or skip through to near the end.)

More cnc milling

John

 

[JasonBModerator @jasonb]

Thanks for the update John. It is actually giving quite a nice finish and not too much in the way of high pitch chatter from the cutter.

It will take a while to get a feel for what depths of cut and feed rates work in various situations on any given machine so keep a note of what works and then you can use the same again or try a bit deeper if you felt it could be taking more off.

[John HinkleyParticipant @johnhinkley26699]

Precisely so, Jason. I’ve ordered a couple of 2mm diameter carbide ball end mills from Rennie Tool Co. which I hope will arrive either tomorrow or Monday as backup.  I think that when I set the Z-probe zero position, it wasn’t sited correctly so it started off diving too deep into the workpiece before I realised and by the time I hit the emergency stop button, the cutter had already broken and pinged off into the enclosure.

I hope to get out into the workshop tomorrow to finish the job and take more videos – as long as I remember to press the record button!

At least my PC seems to be working OK now and the editing suite is behaving.

John

 

[JasonBModerator @jasonb]

Make sure you have the tool ctr set to be outside the boundary. That way it will start taking light cuts as it works its way across. If you don’t do that at beast it will have the ctr on the boundary so the first cut will be 1mm wide.

set outside =0.1

And the first cut

If set on ctr a much bigger first cut

[Mark EasingwoodParticipant @markeasingwood33578]

Well, I am impressed by the results achieved on the Carvera Air

Thank you John for taking the time to film and share your adventures.

The YouTube video below popped up in my feed, it is only 17 minutes long, but does give the feeds and speeds being used to machine 30mm thick aluminium with a 42mm long single flute DLC coated cutter, so it may be worth a look.

Mark.

[John HinkleyParticipant @johnhinkley26699]

Jason,

I’ll do as you have suggested with the extra offset.  Luckily I read this post before I loaded the latest iteration of G-code ready to break another cutter.  Postie is due to deliver the new bits this morning so I’m hopeful that I’ll be successful.

Mark,

That looks an interesting video.  I’ll watch it later when the interminable football is on.

John

 

[John HinkleyParticipant @johnhinkley26699]

Mark,

I didn’t wait for the football, so I watched that video while waiting for the morning post.  I thought that he was pushing the machine more than I would care to until I saw the little “4x” in the bottom right corner. Handy to see the speeds and feeds listed, too.  I wondered how the last operation was carried out with the workpiece apparently just sitting on the wasteboard until the end where the sticky tape was shown.

Jason,

When I checked, the offset that you highlighted was already in place.  I didn’t put it there I don’t think, so perhaps it’s placed by default to help people like me!

John

 

[Julie AnnParticipant @julieann]

Looking good! Nice finish and little to no burrs being raised. That shows that cutter is being worked as it should be. Being cautious and reducing parameters such as speeds and feeds can counter-intuitively make things worse.

Julie

[John HinkleyParticipant @johnhinkley26699]

Yes, Julie.  I didn’t get to do the parallel machining yesterday after getting a phone call mid morning to let me know that a long time friend and work colleague had suddenly passed away.  Left me somewhat shocked as he was a couple of years younger than me and we had been having a cuppa and chat only last week.  Makes you step back and ponder on life and its priorities.

More lawn mowing this morning but I’ve promised myself a session on the mill this afternoon so that will finish the cnc work, I hope. Just tidying up the tabs and hand threading the holes to complete them.

John

 

[John HinkleyParticipant @johnhinkley26699]

I did get out into the workshop and ran the parallel cutting G-code.  I had been too conservative with the number of passes and it took one hour of actual machining to do all four pockets.  This is being edited down to a merciful few minute clips on the video but I’m glad to say that no milling cutters were harmed during those operations.

Here is a still from the video hopefully showing the finish achieved.

As Julie has said, little or no burs except perhaps on the top edges of the long recesses.  I’ll be attacking it next with the Dremel and cutting discs before the final clean up and tapping.

John

 

[John HinkleyParticipant @johnhinkley26699]

At long last, and not before time, I’ve shifted my backside into gear and ventured out into the workshop to finish off the sliding clamps that I started ages ago.  Once you’ve watched the video on YouTube, you’ll be able to spot the machining error that I didn’t see in time to correct on the Carvera Air.  And I was standing right next to it!

The parts could still do with a bit of tidying up with some very fine carborundum paper on a flat surface to pretty them up, but that won’t improve their performance any.

Here’s a still from the video for those that don’t want to watch it:

John

 

[JasonBModerator @jasonb]

Thanks for the update John. Look forward to seeing them in use when you get round to making something else.

[John HinkleyParticipant @johnhinkley26699]

Here you are, Jason, better late than never.  The clamps actually performing useful work.

Spurred on by the recent thread about collaboration with an engine design and production, my interest was rekindled in an internal combustion engine that I started a couple of years ago, but in which I lost interest.  I have been playing around with ideas of a modular engine; that is to say, one which is fundamentally constructed from separate crankcases, cylinders and all the other paraphernalia associated with it but can be simply “ganged” together and elongated from a single cylinder unit into a two- or four-cylinder combination with the necessary extension of the crankshaft, camshaft etc, with the possibility, with suitable design of the crankcase, to be inline, Vee or horizontally opposed.  (I’m thinking perhaps a hexagonal shape to the crankcase together with the appropriate style of crankshaft.)

Every journey starts with a single step,so they say, so I took my original design for a conrod, modified it several times and finally (for now!) decided that a built-up crankshaft was probably the way to go – rather than one made from a single piece of material.  This would allow me to use one-piece conrods instead of my first iteration with split big ends.

Back to Fusion with the design done in Alibre Atom for the production of G-code.  Of course, this didn’t go without mistakes as you can see in the upcoming video (still in production) but I did end up with four connecting rods.

This is a sneak preview of one of them

I’ll be the first to admit that I’m not the most proficient with a file, as witness the marks on the face, but I’m hoping to polish the blemishes out with fine carborundum paper and oil or water.

John

 

[John HinkleyParticipant @johnhinkley26699]

For those interested and with twenty minutes or so to spare, I managed to beat the video editing software into submission – and here’s the result.

John

 

[StevePParticipant @stevepye68246]

Looking good.

I think youve made great progress in mastering your new CNC machine.

Steve p

 

[JasonBModerator @jasonb]

They have come out quite well

You could start the ramp from 1mm above the work which does reduce the amount of air cutting and also have the ramp speed the same as the main cutting speed as you are not going that deep per pass.

The heavy engagement at the end of the contour where the tool moves away into the stock can be avoided if you set the ramp out to zero degrees, have a look back at the clamp thread where I suggested that.

Big ends I would use Pocket to do those which does not leave that spigot in the middle waiting to jam the tool when it breaks loose. I suspect you have a positive number for the bottom height rather than a negative to take you just below or your axial stock to be left was greater than what the bottom height was below the work.

If you are going to put a reasonable size chamfer on parts then you can leave a thin layer of metal all round and not use the tabs, the chamfer will remove the material left around the work.

 

[John HinkleyParticipant @johnhinkley26699]

Steve p,

Thanks for your comment.  Great progress? – I don’t think so -but maybe one step forward …. you know the rest.

Jason,

I agree, Pocket would have been a far better choice for the big ends.  I should have picked that up during simulation runs, but didn’t.  I use Fusion so rarely, it’s almost like having to re-learn it every time.  I set the bottom height to stock bottom -0.2mm.  I thought that would be enough.  Maybe the material is not uniformly thick – unlike me.

Anyway, I bought two pieces of material anticipating a minor catastrophe like this, so I’ve decided to revisit the design, make a few minor tweaks and take up cudgels with Fusion so that I can do it all over again, hopefully with a better result this time.  At least I’ll end up with eight con rods to play with!

Now it becomes clear why, while you have made a whole engine, I end up with four little bits of aluminium!

John

 

[Author]

Posts