Column Stiffener - LMS HiTorque Mini Mill
Вставка
- Опубліковано 21 січ 2018
- Before and after comparison of LMS HiTorque Mini Mill when rigidizing its column with Solid Steel Base and Heavy Wall Structural Tubing.
The Fundamental Frequency was increased from 40-80 Hz to near 100Hz as shown by the HP 3561A Dynamic Signal Analyzer.
Little Machine Shop's cant be beat for the money and weight.
Thank you every one for your feedback, I am humbled that so many experts and hobbyists alike are chiming in! If we are all interested, I could use some help on how to perfect the experimental setup to prove all these great suggestions we are getting (number of flutes, direction of cut, ect…)
THIS video should really be called something like how to excite every resonant frequency in your tool (the opposite of how to produce a good surface finish) - it is not a great example of how to machine, what I hoped to do was push this tool beyond its limits before and after the mod, then try to quantify/identify the thresholds.
I have successfully cut just about every tricky material on this tool: Inconel, titanium, copper, plastic, even ceramics; it is capable of cutting them and making fantastic parts, what it’s not capable of (in any way no matter what you do to it) is be physically stout or massive enough to cleanly remove aluminum at this [in the video] high of a volumetric rate - it did it as you can see but man did it complain - basically the point of this experiment right?
Per specific comments:
I can tell you with absolute certainty that the nonmoving axes were locked, the gibs in the travel axis were tight and the table was not flailing around - further proof is that the gibs were so tight that the whole mill would rather slide rather than the dial turn (rubber pads under the feet got all oily when we moved it for welding); in normal operation there isn't this much force on the dials so the mill doesn't normally slide, especially when the pads are not oily; therefore no need to bolt it down.
Cutter geometry, number of flutes, feeds & spindle speeds.. all these make big differences across comparatively similar situations but to me a good machinist can "feel" around their weaknesses and tune the cut to suit - again, this particular experiment was to push the machine (and tooling) past its limits and excite those resonances so they could be measured.
How do you manage to tram it in with this setup?
@@Sheepisgoo When the column was bolted to the base, I used snap gauges to measure the gap and milled a spacer to suit. Its within a tho or so, you could do the copper and grout trick if you needed, but I don't.
@@joenicotera2991 Yes, the "instrument" will always resonate; I like your association. The objective of a musical instrument is to resonate well over a desired range, and the objective of the string or pipe is to resonate at a discrete frequency (including it's harmonics). Similarly the design objective of a machine is to resonate at only discrete frequencies and with very narrow bandwidth, and be damped over the rest of the spectrum. A milling machine experiences varying loads under different conditions unlike an instrument which ideally your not trying to bend - this is where stiffness helps to minimize the shifting of the resonances (re-tuning of the instrument) over various conditions. As for the horrendous noise, I'm with you its awful, but a good way to excite the machine in the condition in which the stiffness & damping are needed most. If it were done with the hammer under no stress, it can be argued that it wouldn't paint the same picture.
Can you sell a base theres a costumer rigth here
Wow. Actual measurement and empirical science instead of guesswork and handwaving and bluffery. Even clearly stating the limitations of the test environment and equipment. Hats off!
I agree with Ken Ibn Anak in filling the new column stiffener with sand and pea gravel. I have used this technique in barrel tuners (different material of course) for precision rifles and it dampens the flexing of the piece tremendously.
Thanx for the great information, and your time, for putting this together
Glad it was helpful!
Outstanding vid, thank you for the idea for stiffening my mill, I have the same machine, but the older version that tilts at the base, I have never tilted it, and don't plan on it, that join is the problem on mine, and your idea will fix it.
I modified the fine down feed on mine, by making a larger knob/handle, with the z axis tightened up, I can now plunge cut into Titanium without fear of shattering the carbide end mills, well, maybe a little bum tightening ;).
Thanks again,
Scott.
Great idea on the knob! I can imagine myself doing the shuffle trying to twist that thing as I do it - head in the zone, body rotating with the knob...
Filling the column with anchor cement also works very well, weight is everything with mills and lathes
wife must be pissed with the Husky cabinets in here kitchen. The column is a Very nice idea, one of the better ones I have seen.
Thx. Luckily that's the garage :)
@@internationalprecisionengi1737 So she's doubly pissed the new cabinets and counter tops she's been asking for wound up in the garage? lol!
Good science! I love it. I am a loo-thier so I think in resonance. I am not much of a machinist so I watch these videos to gather some knowledge here and there. Here’s a little anecdote. I have a CNC router that I use to make guitar fret boards. The cutter is a 0.025EM. Once I snapped a bit by just barely touching it, that’s how fragile they are. I keep the speed rate very low and the depth of cut very shallow but the spindle speed is something that benefits from this type of analysis. I don’t use fancy equipment but could if I wanted to waste a bunch of time. I determine that the speed of the spindle that is spinning at a resonant frequency can be avoided by a simple method. I put my hand on the spindle and sweep the speed from low to high until I feel vibration. I can mark that speed as the offender. So it’s a simple matter of choosing a speed that is optimal with an optimal feed rate of coarse, And make sure that I am away from that resonance by Turning the speed up or down a little bit. That’s all no big deal but I haven’t broken a bid yet when I touched it by accident.
Yes, I've had conversations with machinists, where the convo abruptly stops and they dive across the shop. "i thought i felt something..." . The other thing that people forget is that the resonances will change as the machine components are tensioned (similar to tuning by changing the tension of a guitar string), so if the machine is loaded or accelerating then it all changes. On a good note, the bits and cutting forces being sooo low on your tool, the resonances should stay relatively unchanged. The next concideration is that wood is organic and doesnt follow equations very well, its individual and changes with the grain... It's all where science meets art!
An interesting and informative video on the shortcomings of this ubiquitous design of mini mill. How about some more conventional experiments just using a DTI strategically placed in various positions between base plate and head that should show the needle fluctuating due to flexing whilst in use.
You could also produce a CAD model and apply Finite Element Analysis to it to see where the greatest stresses arise and so take the guesswork out of a stiffening strategy.
considering the same thing to my g0704 style mill , figured using a U beam thats wider than the original mast and the same type of bottomplate to weld it on but smaller to bolt ontoo a bigger plate the same way the original mast on youre machine , this makes it eazier to machine straight after welding it , i was planning to have the front of the U machined perfectly flat and angled from the ground plate and mount it to the original mast with a few verry big diameter short bolts along its length
as for the frequency of the resonance , seen a verry interesting video a while back from some mayor cnc machine manufacturer , he used the rpm of the machine and reved it up and down in a sinepattern , it preventing the machining to resonate the piece with a large stickout it was machining sounded weird but worked good , because its a cnc machine to can heve the tools move with the same sinespeed to make the cut look even
Yea, that does sound crazy. I've seen chip breaking tool paths too, where the feed is sinusoidal, I think it can load and unload the tool faster that way.
First time trying to cut some pockets with the Harbor freight X2 using a YG-1 3 flute carbide 3/8" and as I entered the corner(double the load if not more) the entire mills "C" shape as you put it folded downward and shattered the end mill right in half. I had previously manual and CNC milled on 250lb - 6,500lb machines but these smaller machines are seriously crippled until you fix the column and by then you still have a tiny cutting area. I paid $450 used like new and I think I paid too much, lol. Now have a 0.5" x 4" x 12" plate bolting column to base in back and I am considering either a new mill or 1" base plate that is 8.5" wide and 12.5" long to bolt the base to then the column plate I made would bolt to the new 1" base as well. From there filling column with epoxy granite and running some threaded rods down into the 1" base from inside the epoxy filled column. Just not sure if 4.3" of Y travel is worth the trouble.
3:56 Your gibs are out of wack. Every crank you make i can see entire jig moving wildly.
Good eye. If you look closely you can see the whole mill sliding on the oily table. I here you, they have to be dialed, and they were. There would be a lot more volatility in the sound if they were loose, the table oscillates at a few hertz, way faster than the dial spins.
Or, you could just bolt it down to a suitable bench like the manual specifically instructs. It is a cool design, and it does what it's supposed to, but I feel you could have achieved similar results simply following the instructions provided with the machine.
These mini mills are top heavy and should never be used without being secured down properly. This is a good alternative if you don't want to run bolts thru your bench, a better option would be to take your design and bolt that to your bench, rigidity on top of rigidity.
Even with all that rigidity you still have to keep in mind that the operative word in "mini mill" is MINI, and adjust your expectations accordingly, lol. I'm giving a thumbs like for sure, very creative.
Thanks mate. Yes you are correct, this mill is maxed out. The previous flex point was the joint and the column, behind the back 2 bolts. The new weak point is the z axis dove tails; you can see the head dancing back and forth. It's almost like someone calculated the loads on the whole machine all at the same time 😄
I think try some materials with bad resonance (cast iron rather that steel). I would try the support beam filling up with lead. Cut some polycarbonate washers to use between support and machine.
I like this video good work
Also I herd about filling it with oil saturated sand, or lead shot & oil. It would be fun to do a more rigorous test but you'd have to straighten the set up out first: the microphone, put a motor on the handle for consistency, get a real carbide bit for rigidity... even then, all that would only push the resonances around not necessarily eliminate them but it would make it much easier to characterize and the pick feeds and speeds!
Excellent instructional video sir! I've read several comments with some sound (no pun intended) ideas, but I think István's is probably the best. The steel structure you added is great for improving stiffness, which is a good thing. ...But steel is also very 'springy' (has a high Young's modulus, or modulus of elasticity) so, by itself, the steel column may not be helping as much as you might have hoped. The added rigidity is good, but adding mass - and especially adding material with a high dampening effect (ie. material with a very low modulus of elasticity, like lead) could do a lot more.
Pouring molten lead into the column would be my first choice. The oil-and-sand idea is good too, though I'd probably use a combination of lead shot, very thick oil (like 90 weight gear oil), and just enough fine sand to fill in most of the space between the shot. The idea behind this is that any vibrations that start in the column are instantly transferred to the oil/sand/shot mixture and the friction induced between the particles & oil will quickly convert the vibration to heat. ...all of which adds up to a basic description of how a dampening effect works.
It's probably self-obvious but, if you go this route, you'll want to make sure all the bolt holes are tightly sealed first, and screwing a gasketed plate to the top of the column could help avert a big mess if it ever tips over.
One more idea I'd like to add would be to bolt (or clamp, at least temporarily, for experimentation) a couple of thick lead slabs to the sides of the mill head, as close to the spindle as possible. Adding 40-50 pounds of lead here might give you a further big advantage in reducing vibrations, especially at low frequencies.
Thanks again for the great video. The data collection and analysis was good, but I really appreciated your inclusion of the how-we-built-it part at the end.
Hi Jacob, I loved this video! Would be great to see how far you could push this idea. Some folks spend a lot of time bulking up their mills with granite bases and such, but you can actually predict, do, and measure to find what truly works, and have the knowledge to explain it. Pretty rare. I've tinkered with accelerometers and have not been impressed with the like from SparkFun. Perhaps one of the industrial vibration transducers is the way to get a clean signal? I don't know if it is appropriate for this work, but I've used a calibrated Dayton EMM-6 microphone (the linearity data can be downloaded given the serial number) with TrueRTA to do audio analysis. With a pre-amp I assume you could bring the signal into your network analyzer too. Best wishes, Kent
Hi Jacob, I thought of this later, but I reading the instructions to a Markforged 3D printer and they have this interesting way to adjust the belt tension that's a bit related to your video. The manual recommends using a piano tuner to measure the tone that the belt generates when plucked and to tension the belt until the tone is near 49 Hz. Thought that was pretty neat.
Thanks bud! The concept can be taken very far, so far that a bunch of people way smarter than me have made a lively hood doing things such as this. One interesting approach is to hit the machine with a calibrated hammer (not kidding) and listen for the resonances with a pricey mic and analyzer. There's no end to the madness, but at the end of the day these are just base points because as a tool wears or changes feeds, the loading changes which will shift the modes... put a bigger part on the tool, changes the stresses in the tool and again changes the modes...
I like the tuner idea, but they should have recommended to play a 50 sine on youtube and compare, just in case someone doesn't have a tuner, particularly a piano tuner. This is cool if you've never experienced it: if you play two tones that are close together you get a beat frequency and as you tune closer to the set point the beat slowly goes away
At IMTS four years ago, one of the coolest demos was this line of seemingly identical parts on a conveyor, tapped by a solenoid driven weight, one by one. A computer recorded their sound and used that to determine if the part had defects. Both my wife and I play piano and use beats to check tuning. Could a non-musician even be told what to listen for? Markforge just recommended downloading a free phone app.
Reinforcing that column is a great idea. I was wondering how stiff the original aluminum column could be.
The original column was as stiff as the dovetails on the Z axis :) we just found the next weakest link. Now that I'm not traveling it is time for the next size up
Try a machinist jack under the other end of the stock hanging out in no-mans land,and clamp it down.
Help to take out chatter and vibration.
Interesting, Ill have to try that. Its definitely flapping in the breeze out there, I wonder how much it is contributing?
Jacob, The usual way to tram the 3960 mill is by using shims between the column and base at the four attachment bolts. With the mill now bolted to the new column/base unit, how are you planning to take care of future tramming issues? If you tram the mill by itself, are you worried that re-attaching it to the reinforcing column/base might introduce enough stresses to pull it out of tram? Did the initial installation change the tramming at all?
Ahh yes the old precision vs stamina debacle. This mill has seen the horizons and beyond and it has considerable wear on the centers of the ways - this makes gibb adjustment especially difficult - so I decided that rather than filling or scraping the ways I will just use it up. I'm amazed at how much abuse this machine can take now that I'm not so concerned about precision: the cuts in the video were 1" x 0.125" at a pretty decent clip - it was like Edward scissor hands at the end of the movie except with metal!
I did however give it a decent effort not to bend the mill column to the new support. We first welded the support and of course that it wasn't square so we bolted the mill to the bottom and used snap gauges to measure the gap and squareness and then machined the spacer to fit. When I put the spacer in it just needed a love tap to get it in, like dialing in a vise. There are some glimpses of us hitting the support with a square and me measuring and milling the spacer at the end of the video (I think).
Thanks for the comment! I appreciate your support, Jacob
Jacob, I didn't mean to diminish your efforts with this upgrade. The 3960 could stand a larger support area on that 4 bolt base, but that would probably require a new base casting. Your approach does offer a much more rigid column, that's for sure. Great video!
I didn't take it that way, it was a great question. My favorite way to tram and stabilize is the Epoxy method. This guy did a great video (link at the end). I made six holes in hopes that one day I would get around to making a 2 inch spacer and raise the column, if this ever happened you could just measure the gap with snap gauges while its bolted to the new column and machine the riser to suit.
ua-cam.com/video/U7Qs-J2swIc/v-deo.html
Its nice to see video's on the Mini Mill. I have one as well but I have a TM motor and 2" column riser. I would have added the column riser before bracing it. I like the design but I believe it prevents you from mounting the airspring inside the column.
I believe you were cutting aluminum, which case if you were, you should have been using a 2 flute end mill. Your issues would go away and the chatter (sound) would as well.
2 Flute has less meat tho so the resonant frequency goes down along with the cutting frequency being cut in half from the number flutes - too much hammering for these little tools, even bridge ports struggle with two flutes in my experience. Its less ideal for sure but it feels better on the dials, give it a try.
@@internationalprecisionengi1737. I agree with Brandon about 2 flutes with aluminum. The difference in the 2 flute vs 4 flute end mill with aluminum isn't about resonant frequency. 2 flute HSS works fine for me on my LMS 3990 on 3/4" 7075 aluminum plate with no chatter.
I have the the LMS 3990 mill with the 2" column riser. I can make heavy side cuts in alu with hardly any chatter (2 flute cutter). I do get some vibration with harder materials. I upgraded the spindle with tapered roller bearings which made a big improvement. I do like the idea of a column brace but are you still able to tram the column?
Thanks for watching and for the comments. 1 inch depth of cut and 0.2" radial even at 2 in/min - that's alot of volumetric material for a weeny mill but it did it. I like the idea of upgrading the bearings too, I'll have to try that.
As for tramming, I didnt bother for the stuff that I do these days. You can still tram it tho, just like normal but rather than having one surface to shim you have two. Kind of like bolting a vice down, set it and snug the bolts, measure, tap it in, tweak the bolts, measure again, tap, tweak, repeat until you send the bolts home!
The column did help but the weak link is the dovetails on the spindle casting, they are too close to the edge, if I wanted to go any further I would plate the sides of the head.
Perhaps having a longer base plate extending to the back would allow a gusset to stiffen the post and thus further dampen the lower freq.
The new weak link is the gibs on the z axis. You can see the head swinging back and forth now, its crazy!
How did you tram the machine after mounting it to the new base?
I can visibly see at the 4 minute mark your table flexing. This means your gibs are probably loose. I filled the column of my own X2 with a mix of sand, small machine nuts and bolts and epoxy up to the bolt hole for the quill control, to reduce resonation, then bolted the table to the mill. Yesterday with a .5" endmill I did the same kind of cuts you show at 4 minutes, in a half inch thick steel plate with no problems. Of course I tighten my gib screws. :) Your high pitch noise is because the material has no coolant, or the speed control is a little high. Try changing the speed of the cutter or change to a 2 flute cutter to change the noise.
All good ideas, Ive always used 4 flute on manual mills to reduce chip load and increase cutting frequencies Ill be less stubborn and give a 2 flute a try again. Usually I run smaller material removal rates and use carbide - I thought this one was carbide at first glance but it was some weird Aluminium nitrite back (purple) coating... fail. The 2Khz singing is the garbage endmill unscrewing, drills are most famous for this. I typically use 4 flute to keep the depth per tooth down for the same rpm, but as you point out the cutting frequency doubles. As for the Gibs, they're snug and the slide lock was half tight. I looked at the video and saw that the motion was the whole mill sliding on the table, prolly because the gibs and lock were too tight. Normally it sits on rubber pads but when we did the welding and when I turned it for the camera cutting oil got under them and rendered them useless for friction.
Wow, 1/2" end mill on 1/2" steel, with how much per pass? Dang, with 0.1" per pass and 4 in/min my mill would be dancing all over - pictures would be falling off the walls in the house... smoke billowing out like cutting fluid is on fire.
Great instructive video! Thanks. I'm surprised how small is the amount of soldering and how it could stand all the strength in a long term.
Have you thought of making several pass of soldering or maybe using a bigger wire later on ? I'm no expert.
That's not solder, it's TIG welding.
Yes it may look anemic, but (if I can remember correctly) the material was beveled prior to welding so they should be full penetration welds; with the correct filler material the weld can be stronger than the bulk even though there is no protrusion.
In this application it doesn't really matter tho, the applied force being a moment on the joint puts only tension or compression on the weld and given that the neutral axis is somewhere between the original column and the new upright, the entire weld will cycle from tension to compression. I haven't done the calcs but I would imagine that even a few tacks would add tremendous stiffness.
@@clintchapman4319 Yes, welding! Sorry for my french accent :) Ok, I thought that the wire thickness needed to be the same dimension as the part to weld. But It's just a support that adds up to the structure I guess...Thanks
Your z axis is dropping as you cut across, the dial is spinning and the quill handle is moving as well. I do like the reinforcement with a separate column and base plate, but alot of chatter comes with not locking the z axis and the table (whichever axis you're not intentionally traveling on, in side milling the x-axis)
Z & X were locked for sure, I can see the position of the Z lock handle based on where I keep it. The worm drive probably was never engaged because I brought it down with the handle. The weakest link is the z axis dove tails, they are right on the edge of the head casting.
At 5:18 and 5:36, your fine feed z axis handle is spinning.
Sorry for being late to the party but what you are describing has no effect. You can see that the main Z axis feed handle is in the coarse feed position (ie. the 3 spoke handle pulled to the outwards position on its shaft) which leaves the fine feed knob completely disengaged and able to spin freely which it likely does from vibration.
few points.. tighten the X axis gibs.. they appear loose and i know these machines get loose in the X over time.. lube is needed at least for your ears.switch to 2 flute cutters for aluminum. Anyway the mill seems a lot more rigid nice solution.. you also need a better Z lock now with the ability to take deeper passes.
Good suggestions. The Z lock is key, the Z dove tails on the head casting are the new weak point. The X was locked and the Y gibs were so tight the whole mill was sliding on the table, if you look close you can see it. I resist normal convention with the 2 flute cutters because: 1. I'm not filling the relief with chips with these light cuts (compared to what the cutters are designed to handle) and 2. a more flutes on the cutter increases cutting frequency (beating induced into the tool by each flute) without increasing surface speed.
@@internationalprecisionengi1737 i dont have the mini anymore.. otherwise id have suggestions for the Z lock.. Mine always got pulled down , even on fairly light cuts the motor could handle no issue.
Do you use raw audio data because audio compression willl ikely eliminate many frequencies.
Whatever the case morerigidity is better, how about some buttressing or stays?
ok, forget the noise. did i miss you recognizing that the largest difference is that the finish on the part is now beautiful. Maybe I did. Forget the frequency
I'm sure this is too late, but I think the mass of the head on this type of mill is a pretty big limiting factor. Now that you've stiffened your column, try adding 30 lbs of lead to the head of the mill. I screwed it on mine, but I think epoxy would work better. The problem with the epoxy is that it would affect resale quite a bit. I really wish someone would redesign these. Why not make the head mount more permanent (though I suppose there would still be some play in the gibs)? Who is going to tilt one of these? Why isn't the column to base mount gusseted? Why doesn't the base extend another 3 inches to the rear so the column could be bolted further away from the column? By the way, if sand is added to the column it will reduce resonance and add mass, but I'd be tempted to add lead shot and epoxy slurry. Sorry I'm so long winded. Stay safe.
Yes, great suggestions. you can watch the head wobble on the Z gibs now. I think you will always be chasing the weakest link. Its better to buy the bigger tool to do bigger work. I just love the footprint on this one and one person can move it! For now I will take more passes, and smaller depth of cuts.
Climb mill instead then conveniential cut the other way to clean up the finish.
Gregg, On manual mills there is usually tons of backlash, as apposed to a CNC where this is minimized by design (one of the reasons for considerable cost increase). Given the backlash, on this machine, there is not enough mass to keep the table from getting periodically sucked into the tool. Even with the gibs sinched, its a balance between the tiny screws overcoming the friction vs the part getting sucked in. When conventional milling you are fighting both the friction and the cutting force so, while not ideal, it is more stable. The optimal algorithm on most manual machines is to take your rough cuts conventionally, and do your finish passes (with significantly less or no load) while climb milling.
What's that slot shaped flat plate attached on the column with a round black disc, right under that black cable hose and above the video timer mark of 5:23? The slot is connected a bolt. What's that used for?
My gosh that sound is annoying at that time! Its a counterbalance spring so the mill head doesn't drop.
@@internationalprecisionengi1737
Thanks for the information. Cheers. 😉👍
Fill the column with epoxy and gravel :)
for sure or just sand and gravel.
I love it, but also mbe bolting it down:) /not to be funny but could a nice rigid stand may aide in absorbing that resonance, No?. Am actually just trying to get a better concept of value (seeing the HFT cost as much as the sieg now!/ & No-coupons). I purchased one & am looking lathes);
Really would appreciate any insight on pricing history, or pre-2021 tariff..
I thoroughly enjoyed the diagnostic process though /& the base is brilliant . Thanks
Yep good point, if I planned more I could get them shorter. I paid like $750 with the precision kit shipped to US ~2010; the mill itself was prolly like $500 USD.
Nice experiment, Jacob. Have you tried recording the audio separately? If you use a good pro mic into a computer running Audacity, you can put that behemoth back in the museum. ;-) Audacity will even do fast Fourier transform, and it’s free. There’s also the FFT mobile app that luthiers use that does fast Fourier analysis.
I’ve read where people have stiffened their mill column with epoxy or even concrete, but I don’t know too much about that. Stefan Gotteswinter did a video where he trammed the column on his mill with machine epoxy.
Your idea to use an accelerometer is interesting. A good pro mic will pick up down to 20Hz and lower, and I’d imagine an accelerometer can measure single-digit frequencies. You said the lowest resonant frequency you measured on your mill was 78Hz I think? Would be interesting to see if there’s anything lower than that going on.
Thanks John, yes you are correct there is lots of room for improvement. From my understanding the pro kit uses an accelerator/vibrometer because of the frequencies of interest and the electrical circuit must be tuned so not to attenuate any signal in that range. As a hobbyist it will likely never get done, I think I'm happy with the take away, stiffness means sharper resonant peaks (higher Q) so they are less likely to be excited.
On this particular mill the weakest link is the dove tails on the z axis, they are on the side of the head. The consequence is significant chatter only along the x axis, so if you put the part in the mill and rough with the x axis rather than the y, it is significantly reduced... so even if we got the hardware and signals dialed in, the analysis would only apply to one single cutting situation. This technique is useful to tune volume production runs where the same toolpath is used for months on end, but it would be extremely expensive to build this data base of situations to be applied across all production - liken it to speeds and feeds selection, there is no replacement for a seasoned salt!
I like the idea of relying on more modern software for analysis, but as with most things software can only get you so far. For the dinosaur, the reason that thing is still getting kicked around is for the analog circuitry inside it. Its amazing, but there are natural components in the filtering that first order math (DSP) just cant handle.
Even the pros use epoxy: Drew Devitt from New Way Air Bearings, he has multiple channels so dont be afraid to do some digging, his stuff is great! ua-cam.com/video/ysyYznf4ris/v-deo.html
Change out your torque spring and put on a air shock on the left side it will lift the head easer
Why didn't you buy a big enough mill in the first place?
That mill can easily swing a 3/4 end mill or a 2" boring head. I've had one for several years. Just yesterday I was milling tool steel in it with a 1/2" carbide endmill. In fact, the one I have has the tiltable column - less rigid than yours. I heard about the "column stiffness" issue and even bought material to help with that. But after going over the mill, it became obvious going through the gibs (tightness and deburring) and properly lubing it makes a huge difference - big enough that the material I bought years ago to stiffen the column is still sitting unused. As with other posters, I can see your table moving (more than the rest of the mill). Loose gibs and machining technique are 90% of your problem, not column stiffness. I'm not saying stiffening the column doesn't help, but in your case that's not the majority of your issues.
Thank you for commenting you triggered a thought, I was wondering why there was so much feedback on this video and I think you nailed it on the head - the video should be called something like how to excite every resonant frequency in your tool (the opposite of how to produce a good surface finish). Measuring frequency and (hoped to but failed at) quantify the magnitude was the basis of the experiment - so it had to be pushed beyond its limits, right? I've tried to summarize some of the major points in every ones feedback and I'll pin it, let me know what you think.
i guess "God forbid" you ever need to tilt it.
It's a solid column so it doesn't tilt. The tilting column is more diverse in utility but the fixed column can take deeper cuts out of the box because its a little more rigid.
Lock your z and x and CLIMB CUT!
Good comments. Yes the axes were locked and the y was tightened. I've never really has success with climb milling on a manual machine, my speculation is that there is too much backlash and the table just jumps around. Climbing works on a CNC, from my understanding, because designers/makers have made attempts to reduce end play in the screws and (almost) zero-backlash ball screws are used.
It unwinds the tool due to conventional milling you should always where possible climb mill
I've never had luck climb milling on manual tools, the table just jumps with all the backlash. Its better on a CNC because they take so much care to avoid backlash. I usually will do spring cuts and finishing passes because as you say the dynamics are better but it only works when the cutting load is less than the force required to move the table.
Trouble is what will your wife say when she gets home …. you've taken all the bench space or good job dear??
I thought it's the garage, not the kitchen.
operator error
Why not simply use an automotive knock sensor? They're just a piezo-electric microphone.
Great idea, I'll check them out. I would imagine that the stiffness of the piezo would be great for high frequencies but would they be sensitive at lower frequencies like 100 hz?
Sorry, but that depth of cut is NOT (1"x1.25"), sure you didn't mean 1"x0.125"
Ha, oops! You would most likely be correct, its only a 1/2" end mill! Good catch
Dude I was so confused on that too until I looked at the comments lol. Too early for this...
I think your rotation is far too low when using aluminum even with HSS tool.. Try less depth and higher t/min. I had the same problem, but by using carbide tools I was over the resonance part of my mill and it was far more comfortable.
for sure, smaller is better with this mill. I normally only do like 0.1 depth of cut, or 0.05 step over at full depth, I was just blasting it to see how it would cry. For the money and size, this mill is the bomb!
парень реально 5000$ за это Г? я за такие деньги смогу собрать на нема34 Closed Loop Servo motor, 3-осевой со шпинделем 4 квт GDK и будет сталь грызть с точностью до 0.03-0.05 мм на приличной подаче. Ход по осям 600х900х300мм.. слов нет, одни маты
The length of time chosen to demonstrate the mill's chatter is like an inescapable, severely awkward conversation
yep sometimes it takes a few revs to get right, sorry...
i believe you are milling in the wrong direction, try climb milling
On manual mills there is usually tons of backlash, as apposed to a CNC where this is minimized by design (one of the reasons for considerable cost increase). Given the backlash, on this machine, there is not enough mass to keep the table from getting periodically sucked into the tool. Even with the gibs sinched, its a balance between the tiny screws overcoming the friction vs the part getting sucked in. When conventional milling you are fighting both the friction and the cutting force so, while not ideal, it is more stable. The optimal algorithm on most manual machines is to take your rough cuts conventionally, and do your finish passes (with significantly less or no load) while climb milling.
Quit using hss endmillls and only use carbide in little mills like this. Yea you're going to save money buying a smaller mill but you'll spend more buying higher quality tooling because I can cut steel with mine, with the tilt column. I don't think these lms solid columns are actually worth it because you still had to stiffen it.
welder knew what he is doing -- mill operator has little clue - watching this waste of time
Thank you come again :)
That milling machine could have done the job. Sounds like you were trying to be too aggressive with how much stock you were trying to take off with each pass. Even though that milling machine is made in communist china, the notorious land of mass-produced junk, the Little Machine Shop manufacture added more quality and performance parts than the other manufactures who use the same castings of that machine for their own versions which their own versions are just painted with a different color, they put their own logo on it, they make them as chintzy as possible, and they and their sellers will hype up the performance of their machines through glamorized advertisements
Good science! I love it. I am a loo-thier so I think in resonance. I am not much of a machinist so I watch these videos to gather some knowledge here and there. Here’s a little anecdote. I have a CNC router that I use to make guitar fret boards. The cutter is a 0.025EM. Once I snapped a bit by just barely touching it, that’s how fragile they are. I keep the speed rate very low and the depth of cut very shallow but the spindle speed is something that benefits from this type of analysis. I don’t use fancy equipment but could if I wanted to waste a bunch of time. I determine that the speed of the spindle that is spinning at a resonant frequency can be avoided by a simple method. I put my hand on the spindle and sweep the speed from low to high until I feel vibration. I can mark that speed as the offender. So it’s a simple matter of choosing a speed that is optimal with an optimal feed rate of coarse, And make sure that I am away from that resonance by Turning the speed up or down a little bit. That’s all no big deal but I haven’t broken a bid yet except when I touched it by accident.
Ken It is incredible how sensitive our bodies are. I worked in a Lab where they did earth quake testing and I wanted to ride the test cell (while OSHA wasn't around of course). I was told that that particular frequency matches our bodies resonant frequency and that it could likely cause serious injury. I never rode it, but always think about it: was it a mental reaction thing or a physical limitation of our bodies to be jerked like that? Regarding your tool: The major benefit that you have is that there is no load while you are milling. This benefits you, otherwise as you tension the members of the machine under load your resonances will likely change. The thing that most people forget is that you can map the best feeds and speeds but you are hitting the entire spectrum of speed when you corner or switch directions (0 to hero).