Improving surface finish by increasing rigidity - Part Two: The Panels
Vložit
- čas přidán 29. 12. 2016
- Help support these videos on Patreon! / chrisdeprisco
Special thanks to James Waples, Daniel Wagner, Jon Einarsson, John Scherer, DOMINIK LIZAK, Ben Knight and Tex© for their support!
I'm attempting to stiffen up the spindly frame in order to reduce vibrations and movement leading to better surface finishes on my parts. - Věda a technologie
![[Part 1] DIY Magnetic Deburrer with Teknic Clearpath Servos and Arduino!](/img/n.gif)
Another fun video, thanks for putting in the work! One thing to look into, crusty old CNC geezer speaking here, is to look into damping. Your whole system looks stiff as a young fella's stick, if you get my drift! What it needs is a bit of something to help settle it down. Now that you made the big box (looks awesome!), try filling it with play sand. Also, if you can box in the column and fill it as well, you might get things damped down a bit. Again, many thanks for putting these videos up, much food for thought!
So very confused, I thought for sure that steel plate was going to be bolted directly to the aluminum base to add rigidity, putting it at the bottom on those thin, hollow legs is going to make things heavier, but that’s about it. Definitely need to add dampening to the frame, pinpoint the flexural points for each axis as well as the backlash in the ballscrew/nut/stepper interfacing. Take a look at some of the granite CNC projects, wish I could remember which guy did one, but it was immaculate, down to microns in flex.
What's the tally on times you've bashed your head on the underside of that table coming to?
nice work.
looks great
Thanks!
awesome progress! a couple of suggestions: to clean up a floor of a part. use a 3 flute EM, single flutes shouldn't finish floor surfaces. Also, If you can, take those parts to a profilometer and get the actual surface deviation if you want to get the empirical data supporting your modification. I love this project and I hope to begin work on my own VMC soon. Looking forward to more videos!
Yea I definitely would use a different EM for finishing; I just love doing aluminum with hose single fluters and high RPM! Good luck, make yours 5x heavier than you think is necessary! :P
+Chris DePrisco Would you mind sharing the details of the cutter, the stock size and the GCode - looks like a nice torture test and it'd be interesting to do comparison runs on the same part you did. Thanks
Southeast Tool SOU323 single flute carbide bit. 24000 RPM, 105 IPM, (40IPM ramp) .375 DOC and .040 WOC; climb milling. The cleanup passes were the same but 50 IPM. Let me know what you come up with!
Have you got a final model of this vmc?
You might consider filling in the interior of the space under your mill with bags of sand from Home Despot or Lowes etc. It worked wonders for a large wood lathe I worked o years ago. Great ay dampening vibration and mass loading a machine structure always helps vibratory problems.
Nice work mate. How many drill bits did you break?
None! Although I did dull up a few. :)
I don't quite get what exactly the issue was. Was it bending or vibration? My CNC sits on rubber pads on a flat table on casters. I don't have any issues with it.
Imagine building a miniature CNC on top of four vertical sticks of uncooked spaghetti. The momentum of the table and vice swinging around would make the table top sway around.
Chris DePrisco I get that but since the CNC is very stiff it should not matter. Is the CNC not screwed down to those giant beams underneath?
That's what I thought originally as well, but apparently it's not as stiff as it looks. It is all bolted together, for sure.
Wow. After seeing those results I'm getting the urge to just make a solid concrete block for my stand. With holes in the bottom for a pallet jack. Is there any reason why that would be inferior to metal?
Jeff;
A 3' x 3' x 3' block of concrete will weigh approx. 2 tons. Make a cage of #5 rebar on 6" centers (embed 3" in from the side walls, top, and bottom) and put two pieces of rectangular steel tubing in the bottom to act as receivers for the pallet jack. Form it up and pour away. Will cost you about $200 in materials. After the concrete cures, you can embed Hilti type anchors in the top to bolt your mill down to.
Don't know if a pallet jack can lift 2 tons, but that will be one solid base.
Regards;
Ron
The height only needs to be 2 feet, so 3x3x2 + 1000lbs cnc would be no problem for my 2.5 ton pallet jack.
I'd be curious to know if Chris considered concrete and ruled it out for any reason.
Maybe concrete is to brittle and it would crumble near the mounting holes. Or maybe it has something to do with thermal expansion not being the same for the metal machine and the concrete base. I guess it could bend the metal.
Or maybe people that decide to build a VMC have the materials and capabilities to easily work with metal and there is no other disadvantage than weight when using concrete. In this case maybe concrete is and ideal solution for DIY when ones abilities and equipment is better suited for concrete work.
Cause this frame was free and I thought it seemed plenty sturdy on it's own. I didn't have an actual plan for a CNC machine and just started building and designing as I went. I've thought about several concrete-ish options for a new machine but this one will have to live with what it is. I think it's a good idea though, let me know if you do it!
Cheap is the main benefit of concrete, so free metal is tough to beat :) You might just have the highest quality to cost ratio of any CNC machine ever built here, and it's been great to watch it come to life.
Another idea: Get the biggest sealed container that you can fit on the bottom shelf of your table and fill it completely with water, sand, or both. Just more free-ish mass to help deaden vibrations.
Chris;
Have you checked your mill for tram in both the X and Y axis? I know you are using a single flute mill, which is not great for surface finish, but it looks like the mill might be slightly out of tram.
It also sounded like you had a harmonic going during the cut and my experience is that usually indicates the mill is chattering. The side profile in your hole cut looked like it was showing a finish consistent with chattering, even if minor. It's funny how minor tweaks to RPM and feed speed can make big differences in the finish of the piece.
Great work as always and looking forward to some of the projects you are going to come up with.
Regards;
Ron
Those calipers are done!
Looks better but try Onsrud super "O" single flute endmill, 65 series. Have a sweeper at the bottom, which makes the pocket floors smoother...
wow, that extension cord. Cheaper than a plug, right.
I have a dedicated 50A plug and breaker for the welder but as you can see, the cord is very short. That's normally not a problem because I can move what needs welded. This time I couldn't and I wasn't near the plug. I wasn't about to go spend ~$100 to make an extension cord for 60 seconds of welding.
Chris DePrisco makes sense, yeah. I don't want to sound like a mother, but stay safe, we wanna see more of that machine.
Chris DePrisco When I was in that position (needing to weld out in the driveway) I went to Home Depot and looked through their offcuts and returned stuff hanging near the bulk wire dispenser rack thing. I found an 8 gauge 3+ground stranded cable like 20 or 25 feet long. I tracked down the guy who worked there and asked him how much. He looked confused and said they don't sell 8 gauge stranded. He scrolled through the entire rack to confirm. Said it was probably left by the a contractor doing work on the building and someone assumed it was inventory. SO HE GAVE IT TO ME. Even walked me out the door so no one would hassle me.
Of course then I had to go back in and spend $40+ on connectors for both ends to turn it into an extension cord. Those things are so expensive!!
I think self tapping screws would have been just fine to hold the panels on and would have saved you a lot of drilling and tapping
I think that may have been a good idea...
I still think you should brace those 3 spindle brackets
Me too. It's on the to-do list. I keep hoping I'll find a "real" spindle to replace this one with; at least that's the plan.
that would certainly be the better way, i'd also maybe go as far as using steel instead of aluminum for the plate and brackets (with the benefit of added weight).
Do you have a second backlash nut for the vertical ballscrew? (maybe something to also consider since it seems that you get vertical movement based on the surface finish, just a thought).
Chris DePrisco It is hard to say for certain but by eye I swear I see axial runout of the spindle as you helix that hole. The step at the top of the rough bore and the floor finish having chatter seems to validate it. As insufficient as it might be I'd suggest cutting down the ramping angles until you can get a more rigid spindle. Also there are end mills that have one flute that on the bottom have a wiping effect as it cuts across the floor/face which could greatly improve your floor finish on aluminum.
I will try and measure the axial play tonight and report back. Definitely there are better options for surface finish; I just really like using that single flute, lol. Too bad I trashed the cutting edge hitting that dowel pin... brand new end mill, too...
So good news: there is virtually no axial runout that I can measure in the spindle. With a lever and a load there is the same vertical movement (~0.002") measuring from the table to the wrench flat on the spindle as there is measuring from the table to the aluminum brackets holding the spindle. i.e. both the bottom of the spindle shaft and the aluminum brackets move the same amount, together.
From the table to the Z carriage at the base of the linear rails there is 0.0005" or less (so that's not moving); and from the base of the Z axis (definitely not moving) to the carriage at the base of the linear rails there is the same, approx 0.0005" (so the table isn't flexing downward)..
In other words, it looks like any axial movement is coming from the shoddy way the spindle is mounted, not in the spindle or Z axis upright assembly - like sharp x said.
Why weld the frame? It will pull out of alignment? I've seen 10mm steel plate structures pulled out of square because of one tig bead and these tubes seem pretty thin.
I think it will be fine since there is a whole lot of steel right there that is already fixed in place. Just a few small beads shouldn't move it enough for me to be worried about. If course I could be wrong; I think that's happened once or twice. :)
Look Chris, I'm and expert on the internet and I know best. Alright?
All jokes aside - the machine looks absolutely great and the difference in surface finish was easily noticeable. Keep improving man :)
Btw - are you doing this as a day job?
Thanks! It will just keep improving as I keep finding weak spots and patching them up as best as I can. I wish this was my day job... maybe one day...
That's an improvement did you here the ringing when you screwed the panels down? Rubber mat bonded/ Hint. LOL tractor supply barn floor mats? Hint
Weakest link is now the four leveling bolts.... Must have lateral deflexion now.... Maybe larger bolts ?
maybe do a modal analysis and which eigenmodes move either the work or the spindle. Then you can test out improvements in you computer model before implementing
Have you checked the run out on your spindle ? Seems like people with spindles like your have had great success with grinding the taper and getting a lot less runout. Just a thought, may help surface finish as well .
I have, it's about 2-3 tenths. Not sure how I'd make that better with the equipment I have.
Thats better than most. Some people were seeing 2-3 thou. There are vids on youtube of how to do it but it probably wouldnt be much better than you currently have.
Wow, that's nuts! This is my second spindle from ebay and the first was actually even better. It's always a gamble there though, for sure.
Chris DePrisco I highly recommend this book, I have it in German, and really want it in English as well, but that will have to wait for another day. It has all of the speeds, feeds, depths of everything you can think of, and more.
[(Mechanical and Metal Trades Handbook)] [Author: Ulrich Fischer] published on (January, 2013) www.amazon.com/dp/B00Y4QKF3S/ref=cm_sw_r_cp_apa_FlYjzbP2ZCT80
It's a direct translation from German to English, and I do suppose it's in metric for the most part.
Please include subtitles in the video settings.
noice
you should definitely get a 2nd build going with a proper frame for your machine using epoxy granite to form it from a mold and embedded plates for rails and such
I wish I had the space and money for that!
4 pieces of sheetrock cost less than $30. just saying.
The spindle is half your problem the rest is the frame
That spindle is known for lack of balance and runout with pro high speed machining the spindles cost a fortune for the 2 reasons above.
I highly recommend dropping to bellow 10,000 rpm servo spindle (high nm under 2000 rpm then it starts to drop)
The frame needs to have some form of vibration dampening machines are made from castings due to its high dampening characteristics then its real granite and top of them all is mineral casting (10 times better then casting) believe it or not you can achieve mineral casting standards
You will need quarts and a west coast type resin. You can start filling all voids with the mixture especially z axis and table.
Vibrations should be drastically reduced before either point reaching the collum.
I hope this helps and thank you for what you do and your impeccable taste in music on other vids
look at your spindle itself in the machining footage. you can literally see it visibly wobble. need a more rigid spindle
I think its coming from backlash...
That could be, although the plates did improve things quite a bit on their own!
If you'd like to do math ill give you the speeds and feeds for what ever metals/ router metals.
Still terrible finish.
I wonder if it's that aluminum base plate. It's long length to thickness is probably contributing some harmonics that are amplifying vibrations that start from poor rigidity.
Also, you've hung your spindle way too far off of its base, with a poor design. So you have poor rigidity (again) and a long lever effect. Try a redesign of your spindle mount. Move it as close to the z rails as possible to reduce the leverage, and box it in with thick plates on the sides to increase rigidity.
You whipper snappers and your rocket science electrickery know how just blow an old retired steel fabricator away. that being said like the old gearheads used to say "there's no replacement for displacement" when it comes to machine building that translates to there is no replacement for mass. If your base was the same size tube as the 2x4 or whatever it is under your machine but with a 1/4" or 3/8" wall with panels bolted ot it
A little narration would improve the videos. Even if stating the obvious it would entertain rather than seeing a color 1910 movie.
Never do a plug like that!!
How come?