Measure & Fix Tool Runout! WW113
Вставка
- Опубліковано 20 вер 2024
- We're going to machine an injection mold using very small end mills! Measuring and adjusting tool runout is important to maximize tool life, avoid tool breakage and get good surface finishes!
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Links for this video
Noga Mount: amzn.to/2d2ayv5
Mitutoyo DTI: amzn.to/2dsePri
Camera Equipment:
Canon Body: amzn.to/2cE1765
Canon Lens: amzn.to/2d21iZH
Sony: amzn.to/2d22kEL
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Reach us / CNC Info:
Speeds & Feeds: provencut.com
Download Fusion 360: www.dpbolvw.ne...
Online Fusion 360 Training: bit.ly/LearnFus...
Hands-On CNC Classes: www.nyccnc.com...
SMW Products: saundersmachin...
CNC Resources: www.nyccnc.com
Music copyrighted by John Saunders 5 Reasons to Use a Fixture Plate on Your CNC Machine: bit.ly/3sNA4uH
Finally a great vid on runout!!! This will be helpful for many people.
Spent an hour at work today reading about runout without noticing this video was in my notifications! Awesome job looking forward to seeing the mold.
you could make yourself nuts checking a tool like that. you also need to know whether you are dealing with axial or radial runout, or a combination of both. i would start by checking the tool holding system .put a gage pin in the collet and check runout .if you cant eliminate runout there you are wasting your time checking a tool. i would think the tool shank would be the most accurate place to check runout since the tool is ground relative to the shank. i would also make a test cut with the tool and then check runout. it may settle down. if the tool is running out would it not cut an oversize slot? this might be a quicker way to test it .i'm glad it you and not me. i'm interested to see how you solve this .good luck john and thanks for the vidios.
I am a machinist by trade and here is what I have found out. Trying to dial in an endmill or reamer, I have even seen people do this with drills, by tapping on the shank of the tool is a waste of time. Once the tool takes a cutting load it will return to its most natural state, which may result in a poor cutting condition. I also do not check run out on tools before I run them. For me, it shortens my set up time to not check every tool in a collet and I can get the job rolling faster. If problems do come up that are related to a run out issue on a certain tool, I then start investigating the issue. Using an indicator set up as shown in the video I check the run out. If it is not in an acceptable range I start by rotating the tool in the collet and the collet in the holder as quick checks to see if any change occurs. I have gotten better numbers on occasion this way. Next and most often thing I do is switch out the collet. More often than not excessive runout in collets is an issue with the collet itself being worn out or manufactured sloppily with lax inspection methods. It can also be due to a worn out collet cap but I have not seen too many that needed to be replaced. Keep them clean and torque them to their spec so as to achieve consistency. Even more, do not stuff a tool that is just a bit to big but still fits in a collet, only smaller tools. Check for your specific collet compression, we use a tool no smaller than 1/32 under the collet size.
Really good video man but i have to ask about the way your using the indicator on the tool. Most indicators come with a sheet that tells you if your indicator is at an angle not perpendicular to the part being measured you have to calculate the actual error using the reading you get and the angle of the indicator tip to get the real error. I have not tried this myself because for what i do its trivial but its probably worth noting when using such small tools. Let me know what you think im not an expert i just love learning as much as i can. Im self taught and plan to open a shop of my own one day so we have that in common lol
If you can alter your runout just by repositioning the toolholder in the spindle, do a good cleaning and check the inside of the spindle. It's one of the most neglected things in a shop, probably because it's rarely seen.
But dirt, grease, dried oil/coolant, small metal chips are all culprits for not only run out, but for causing irreparable damage to the spindle bore.
As for ER collects they about middle of the road for runout. Hydraulic and shrink fit being the most accurate and sidelock/Weldon holders being the worst.
It depends on the application. though. Roughing, side locks are one of the best in taking lateral forces. They're better than even hydraulic holders for that. And far better than shrink fit.
ER are for light duty milling and drilling. They have accurate runout and good axial stability.
Biggest thing is clean. Clean clean clean. Tapers, collets, everything. Doesn't take much surface oxidation to kick a collect taper out a few tenths. Also make sure the slits in the collet it's self are cleaned out. Tiny bits of metal like to get stuck in them and get packed in and caked together by drying coolant.
I am " super excited " for this wednesday.
Best of luck !
Nice multiple video screens John. Really gives a good perspective.
John, I have a couple observations. 1, Measuring on a flute at one vertical position, only shows runout at that point. Since it is a spiral flute, it may still have runout at a different height which will be in a different rotational axis position. 2, It may be easier to tweak 0 by using the .0001 jog wheel on your cnc instead of a fine tune wheel on your magnetic base which many people will not have. And 3, I have bent the needle on a thankfully cheap Shars brand indicator by tapping a part true while indicating in a lathe. I knew not to tap in the direction of the contact point, but I didn't consider the rebound I would get from the impact would be enough inertia to bend the needle. That was a .001 indicator. I bet a .0001 will be 10 times more susceptible to damage.
Why this 'tapping' works:
Imagine the collet as a bunch of "blades" pressing on the shank of the tool. The blade are slightly flexible. Because of that the "blades" can be twisted and the tool itself can move around. The friction between the tool and the collet and the collet holder prevent the readjusting if the force es are not to big.
You get the run out in the first place because the collets are not perfectly symmetrical. So some part is putting slightly more pressure on the tool than other. Maybe one "blade" is slightly twisted or shorter or wider. This is not such a problem for bigger tools, the "blades" pressing on the tool are more rugged and the contacting surfaces are bigger, equalizing inaccuracies.
So you can get no run out with a big tool but big turnout with a small one.
Not sure if it was, but when your are down to measuring 10ths you should have your spindle bearings upto temp, the same as surface grinding. 🖒
really looking forward to seeing this finished mold as well as the before/after runout comparison, I hope we get to see some of the fusion work!
Great Video John. We use a lot of Shrink Fit holders. But you can only use them with solid Carbide tools (not insert tools). Otherwise we are using TG-100; Face Mill Holders; End Mill Holders; Milling Chucks; ER-32; ER-16. We have CAT-50 Spindles on our LeBlond Makino machines (vertical & horizontal), and CAT-40 for our Haas VF-3 & VM-3.
Thanks John, I learned a lot. Looking forward to seeing you cut that mold.
For larger end mills and reamers the perfect tool I have found to nudge tools into concentricity is a bar of particle board with a concave curve on the end contacting the tool. No hammer needed for tools under an inch in diameter. It is soft enough to dampen vibration and keep flutes sharp.
I took and engraver to my spindle and made a indexing mark on it. When I needed a tool, like my probe, to have zero run out I would spin it until it was zero and then mark the tool.
Tapping a tool will only help 'seat' it in the collet but probably won't fix all your runout.
ER collets do have a lifespan, but also the holder. Holders will wear just as much as collets and need to be replaced almost as often in sum applications. I find that I will buy a new quality holder and collect for a tool that requires very low runout and if there are no tool breakages (instant collet destructo) I will get many dozens of tool changes out of it without a problem.
One of the biggest enemy to collects is distortion from over/under tightening. You always want to tighten a collet of any type to the spec, no more or no less, under tightening will not fully seat the collect in the body even if it is holding the tool enough and this can cause fretting. Over tightening is also if not worse bad because it will distort the body which in turn distorts the collet. This can lead to deformation of the collet and holder. This will forever ruin both the hold power and accuracy of a collect system. If you see someone hammering on a collet wrench to get it a tighter so that it will hold better they are actually reducing the holding power and destroying the holder. If you don't have a feel for what tight is then go spend the money on a torque wrench as that is cheaper than replacement holders. Torque specs very quite a bit from system to system and depends on the style of nut, plain or bearinged. You will need to look up the exact spec for your collect size and system.
Another big Issue that gets over look is not having enough shank in the holder you need to cover the entire bearing area of the collet with the shank. If you don't the back of the collet, or the nose if you got the tool to far in the collet, will collapse more and distort the ID of the collect and most likely damage it.
Speaking of damage, If you break a tool shank inside a collet most likely its nose is damaged and probably is junk. If the tool shank breaks completely outside the collet you may be okay but I usually mark these for the roughing tools only. More often than not there in a bad crash the is damage to the holder and it also should be downgraded.
Lastly cleanliness is next to godliness! Clean everything twice and wipe all the mating surfaces with you finger, not a rag! rags and towels can leave lint behind and your finger will feel damage and grit that you can't even see. Don't be tempted to just loose off a collect enough to remove and replace a tool, take it apart and clean it. Dirt in the mating surfaces will kill your holder and collets life and cause runout.
Hurrah!!! 2 tenths!!! Dead nuts!!! Almost we called it broken indicator!!!
Grinding specs from manufactures are different for different parts of a tool at times. True run out can best be measured only on primary cutting edge.
Edge mills have a primary and secondary relief on cutting edge.
Don't forget the accuracy of your spindle itself and how much play there is in your ways and wear and tear on the collet in your spindle. FWIW, I spent several years running almost nothing but tools smaller than what you're using here.
When the endmill is made, they may not worry that much about the tapered area, so probably not a great place to measure. The shank and the flutes should be referenced to each other. Also what is the basic accuracy of the Tormach spindle, and will it repeat after running? When chasing tenths, everything matters and metal seems to be made of rubber.
Very interesting John, Looking forward to the machining!
your vids are extremely helpful for an apprentice machinist like myself ... thankyou John !
+NYC CNC very welcome ..keep up the awesome work !
im gonna order a set of those kant twists ... is there any way you could engrave my initials in them at the same time you engrave your logo ?
If it moves that easy in the collet would just running an edge cut straighten itself out by pounding higher on the high side. Be interesting to see if the cutting forces really are lower than those hammer taps.
Dont you think it self centers in action?
Great video John! Hope this "little project" isn't kicking your butt! Looks like your going along smoothly! Just wondering, do you think you can get away with getting those curved surfaces with an end mill? Or are you coming back with a 1/32 ball mill? Great video once again! Hope your learning a lot because I sure am!
You should turn up a TTS shrink fit holder. If you mill-turned it, I'm betting it would have less run out then the ER collet holders.
John,
Are you aware that there are adjustable ER collet nuts that allow you to dial out the run-out?
Great topic, thanks for careful coverage and discussion. I like to see chips flying (perhaps like others!) but setup and gauging are important. A bit dry maybe but can make the difference between CAD perfection and the reality of a part in your hand.
Something else to consider is the quality of the ER nut.
I have some cheap Chinese CAT40 - ER 16 holders, you know the ones that are 3 for $90 on ebay. The inner barrel shows 0.0000" runout in the mill but when I installed my test bar I was getting 0.0010" runout.
Messed around changing collets, moving this, moving that and finally used a Maritool collet nut. BOOM. test bar shows 0.0000" runout.
Interesting technique with the screwdriver. Have to give that a try next time.
WOW! Excellent info John. Ciao, Marco.
OMG man, i feel like i am lost.
I didnt expect those runouts at all... well, how i suppose to reduce the runout in a personal cnc ?!!
any tips guys..
Thanks
When I indicate lathe tools like steady rest in lower revolver and sometimes the work piece I always lift the indicator off before hammering. The pin moves if I don't take it off and then it's just more hammering.
It doesn't look like comments/questions are answered here, but I will ask anyways. Are there questions of accuracy when you are changing where you are measuring the runout? I would be doubting my results if I measured .002 on the flutes but move to the tapered portion and measure .005. I also am a complete newbie to this and am trying to learn as much as I can before I start (non-Tormach CNC coming in the next month or so).
Did you check the runout after a milling operation to see if it held?
another super informative video..thank you John
yeah i've worked at places where they tell you to get the run out as low as you can in 5 minutes and it would be programmed to a certain number before it runs a finish pass and just inspect the number were your machine is actually cutting on and on the finish program we'll have cutter com running the pass with the shift in it.
you can also touch the bit on ground part and run it back and forth on x axis to find your high point then zero it out on the dial.
This week I used a milling cutter with a diameter of 0.2 mm. the length of the cutting edges is 0.7 mm. Preferred speed for 52 hrc steel 1.2344 is 420 000RPM when doing 3d toolpaths :)
Could the cutting section be bent (in relation to the taper and shank)? A measurement at the tip and close to the taper might tell, although the peak of the flute will be at a different point on the "clock".
A light cut using the full length of the flutes might show whether the tip was bent away from the axis (by cutting deeper near the tip). Everything else (work holding, tram, etc.) would have to be MORE accurate than what you want to measure. A few tenths angle error over 1/4 inch (length of cutting) might be tough to achieve.
Wow. Great tip! I really would like to know what is exactly going on when you beat the tool in like that. It must be maybe the collet slipping around in the inner taper? That is my best guess, I'm pretty sure that the carbide could not yield, if it is deflecting I assume it is bouncing back. Make sure to check the TIR of the tool after you cut the part, that would be interesting to know. Thanks for the great vids!
Shouldn't you be checking the runout with a precision ground test bar?
0:46 vocally, you said that right, as what people would say in the shop. You have the "Thou", and then you have the "Tenth". Which is for 1/10th of a Thou. That makes sense to people doesn't it I thought? No one says Tenth to literally mean 0.1 of an Inch. I see that number as 100 Thou. This is shop talk. (in America anyways) :)
You should have put a indicator on the collet holder while hitting the tool with a hammer to check that it wasn't just moving tool holder within the machine.
Well done video on a great topic. But, if the shaaft isn't concentric, that being the part of the tool the collet "grabs", then the tool will never be concentric - am I missing something?
What you're saying doesn't make sense - maybe you just phrased it wrong ;-) The shank of the tool should be completely straight and concentric with the "business-end" of the tool. If it isn't then there's something wrong with the tool :-)
usually you also measure runout at the tool shaft, isn't it? - I mean the question isn't if the tool is okay and the shafts are also grinded so...
I thought he meant the taper, not the shank. But I couldn't quite see in the video where he measured.
If your chucking off the shank all your indicator readings are coming off of that, those shanks are precision ground.
What about the table run out? was the table locked? Better to place mag base on a fixed point on the mill what does not move. Good video, and first wouldn't you want to check machine run out before tool run out? thanks for the video, Greg
Which part of a mill has more runout? The tool/ tool holder / spindle or the table? I can see where setting up the tool to very close tolerance could all be for not because of table differences moving in different directions on the x or y planes.
Interesting topic John, thanks for sharing... Fred
with the finger behind the tool it'll help you with control and less likely to bounce back. same concept at a dead blow. if that makes any sense.
whats your tolerance on the surface profile? knowing that tolerance should give you a good reference on how tight your run out should be. anywhere from 0 to 4 tenths should be ok unless its calling for a crazy tolerance which I've seen.
John, Would you say that using a gauge pin could also help to identify the problem? i.e. no run out with gauge pin, problem lies in end mill; run out, try a different collet, if run out goes away, problem resided in collet, if run out continues, try indexing tool holder, if it follows indexing, problem is possibly in tool holder/tram/spindle?
One question I've always had is the following: lets say I have a machine of similar quality to the Tormach, it's well aligned and run out in the order of 0.01mm. What accuracy can I expect right of the bat? Say, if I program a round pocket of 50mm diameter, what tolerances can I expect?
I'm wondering what the Tormach 440's spindle run out is. I also wonder what that Tormach tooling's run out is. It appears as though their tooling doesn't locate on a taper, leading to even more run out. I try to avoid using collets to hold end mills as they're not as rigid and induce another stack of runout in the assembly.
How can you check runout when you are not attached to the spindle, seems like you might also be moving the Z axis.
Thinking about it, when you get that small, .0005 runout is 15%.
Just confirming, the tool is tightened down when you're hitting it with the screwdriver?
Thanks John. One question, where do you get your micro drills and moreso, what collets/tool holders are you using for them? I appreciate any info!
Did this endlessly on a Moore Jig Grinder with CBN grinding wheels with carbide shanks, would keep on rotating in precision collect and still would tap in with wrench and indicator .ooo1"
hey when i do this i just use the table to bring it up to the tool! Now i don't know them mills can do? what kind of numbers can you hold? I work in a mold shop and i have to hold numbers within +-.0001 all day and sometimes as much as +-.00005. It is a Whole new game and to us Tool Run out can be the difference between a good part and a bad one.
so with such a small cutter, doesn't it go out again as soon as you start putting a side force on it when using it to cut again?
yep
How you put the runout of endmill into fusion?
The measuring force (0.3N = 0,07lbf) of that indicator is so small that it won't deflect the end mill in any reasonable amount.
"Such a huge tip for a small tool" hahaha
Shouldn't it matter if the dial indicator isn't in plane of the axis you are measuring for runout? I know you are concerned more about the delta than an actual true measurement but putting the indicator in plane should minimize potential deflection on the tool being measured.
As long as the indicator is around the correct angle to measuring surface no cosign error will occur. Most companies list how to deal with this in their information you get with the indicator.
Are there maybe single flute endmills in this size? They should have no real runout.
Klaufmann
Single flutes can still have run out within the tool holder or the spindle. So if it's out of concentricity with the spindle it will cut oversized. You just won't have to worry about the runout affecting the tools chipload.
Like Occams says this will affect the chipload, so you won't get uneven wear using a single flute you'd probably get a better finish (when comparing both to each other with zero runout). However, just being single flute doesn't prevent the tool from being off center making the cutting seem like it was made with a larger diameter tool or being set at an angle like "/" both of which put unwanted stress on the tool and poorer finish.
Yeah, that's what I meant.
The indicator should be perpendicular to surface being measured. you have cosine error making your numbers look bigger than they actually are, but I also understand you are only concerned manipulating the difference to zero, not necessarily getting an accurate reading
Wow,about thousandth out to start!
Hey john, just a thought. When you first checked the run out (0.0002) you were closer to the end of the end mill. After checking the taper you than thought you would double check the flutes, I did notice the the position of the stylus was a lot higher the 2nd time. Would it be fare to say you were measuring the same run out but at a different axial position due to the helix being say 90 deg from the first position on the end??? From the end mills perspective - Run out x verses run out on y?? Hope that makes sense. Dave
That is the same I am thinking.
At one Z position you can check runout only in one axis. This means that should check runout on two positions of the Z where flutes of the helix are 90 deg from one check to another.
Correct. His run out could be angular "/", not just off center "| ". There are collets that allow you to adjust both
I suppose my thinking is more so around how it is measured, not so much wether it is out concentrically or angularly.
Have you tried using a hydraulic chuck? or putting the dial on the spindle taper itself?
There is no hydraulic chuck for the Tormach machines.
kinda funny, i stick an endmill into a SK16 holder from maritool with their collets and its never more than .0002"
Hey john im ur number one followers
You saw that collet tapping on ROBRENZ channel.
ROBRENZ ROCKS!!!
John you think the indicator pressure on tool could make it run out??
Thanks
wouldn't that even out itself?
+Klaufmann
Good point!
if you have to tool change like this everytime.....
Most times I've had run out problems was with the Amana insert bits. I've wasted hundreds on them and won't buy again. Sorry but after my continued disappointment I've switched back to Carbi-tool yet again. I won't be buying those over priced bits from tools today again. 😔
To me this sounds like peening that was explained by Stefan Gotteswinter not long ago: ua-cam.com/video/5QE-1S-3HQY/v-deo.html(around 4:30). But I wouldn't do that in the spindle. The slightest knock leaves marks in the bearing surfaces: ua-cam.com/video/Pmplff_f5pI/v-deo.html
.0005 out... I don't see the problem. fix it with cdc. Don't sweat the small stuff
what is tool runout?
watch me wail on my lathe with a hammer next. Lol, I smash high spots off that the exact same way.
hey ..
use EDM which is more convenient for this type of work...