what cause's coining aka dwell marks?

[Delw]

I have one machine that has gotten pretty bad over the last 1.5 weeks. every time it changes directions it leaves a coin/dwell mark. what causes this. I was thinking it was thrust bearings, a tech mentioned drawbar issues.

this part is roughed leaving .008 for a finish pass, the part pictured is after the finish pass. we ran these exact parts and exact program 4 weeks ago with no issues had few very small marks maybe a few tenths deep. tumbler took them right out. the new job we started last week, 1st 2 days no issues then it got much worse as the week went on. so ended up reprogramming the part so the endmill changes dir off the part
went to the second side today its getting worse about .002+ deep dwell marks.

the machine repeats all day long all axis's.
popped a indicator have about .0005 play when I move head with alot of force by hand, tried just moving tool in holder and the same .0005. (waiting on tech to stop by and check draw bar force.

that all being said a new endmill was used as well as the old endmill 1/2 s-carb sharp corner made no different. I dotn get the coining on the 1/4" or 1/8 endmills. Material is alum 6061 .
changed the program so it runs a rad on every axis change and slowed up the speed its more visual now maybe a few tenths deep at the most. while that does get the part working it doesnt fix the machine, I'm sure its mechanical.
maybe could be feed back bleed on the z axis motor when the x and y are moving??

the older fadal doesnt do it and the hass does it once in a blue moon but its only visual not more than a tenth id that.

 

[Vancbiker]

I've seen similar digs from bad spindle bearings, but not related to direction changes. Since it is happening at direction changes it could be a leveling problem. Also, what machine is this? I have seen traveling head type VMCs (Mori Seiki M300 series) do this from worn linear ways. My gut feel is leveling since it was not as bad just a month ago.

 

[CNC_Chip_Thin]

Tough question as it could be a range of things. One would think there is a connection with the fact that this just started happening.

It does look like this is occurring when your toolpath makes a 90 degree turn in any direction. So first places I would be looking are...

Is there a backlash issue that is making the tool dwell for a second and possibly pulling the part up or tool down?
Are either the roughing or finishing tool dulling out causing more chatter, part pull up or tool pull down?
Or have there been any other significant changes in the past week and a half that would cause the change?

Probably just stating the obvious but that's usually a good place to start before really digging into possible solutions...

*On Edit* Vancbiker beat me to the post, at least I know I'm of the same thought process as other's here

 

[Mud]

Warning - I'm gonna blurt out the first thing I thought of, from a completely different machine. Hope that's OK here.
I had a Cincinnati 10VC (2 actually) that made marks like that at seemingly random points. It did it less during first run parts with rapid override on. I finally traced it to the toolchanger operating. The head had a hydraulic counterbalance cylinder, and the toolchanger was hydraulic, when a TC solenoid valve opened or closed it changed the counterbalance pressure slightly and moved the head just enough to make a mark. When rapid override was on, the tool prestage movement was complete before it started to cut. Usually. So the only fix was to not prestage tools during crucial finishing cuts.
If I was visiting, I'd grab the head while it was cutting and see if I could shake it and make marks, just to look for movement as a start.
What kind of counterbalance does that machine have? Weights on chains? Do they operate smoothly? Anything disturbing them?
An air cylinder? Is it working right?

 

[Delw]

you guys make some good points. because of those points I have noticed a few things that could be the problem and never thought of it.

I've seen similar digs from bad spindle bearings, but not related to direction changes. Since it is happening at direction changes it could be a leveling problem. Also, what machine is this? I have seen traveling head type VMCs (Mori Seiki M300 series) do this from worn linear ways. My gut feel is leveling since it was not as bad just a month ago.

Its on a 2000 30x20 box way fadal. this machine almost twice as big casting wise as the standard 4020 fadal. its called an extra rigid machine or something like that.

your leveling idea makes sense as I havent checked level in quit sometime. we got it a year ago, been running everyday since then. will check it in the AM.

 

[Delw]

Tough question as it could be a range of things. One would think there is a connection with the fact that this just started happening.

It does look like this is occurring when your toolpath makes a 90 degree turn in any direction. So first places I would be looking are...

Is there a backlash issue that is making the tool dwell for a second and possibly pulling the part up or tool down?
Are either the roughing or finishing tool dulling out causing more chatter, part pull up or tool pull down?
Or have there been any other significant changes in the past week and a half that would cause the change?

Probably just stating the obvious but that's usually a good place to start before really digging into possible solutions...

*On Edit* Vancbiker beat me to the post, at least I know I'm of the same thought process as other's here

Yes all 90 degree, fadals tend to dwell for a micro second even using an M8 there control is kinda slow.
the last job I ran 145 pcs with one 1/2" endmill roughing and finishing and no issues, this job I tried a brand new endmill had no changes.
no changes in the last 4 weeks the machine runs everyday even have the same tools/tool holders in it as last time.
today when I tried a new endmill I also used a new holder and pull stud. thinking the pull stud wore(weve all seen the little marks on the underside of the angle). Ive been using the same pull studs for the last 20 years on the fadal and had just gotten a new batch of them last week.
I will try to move the head while running and see if I can mark a part up. good idea. thanks

 

[Delw]

Warning - I'm gonna blurt out the first thing I thought of, from a completely different machine. Hope that's OK here.
I had a Cincinnati 10VC (2 actually) that made marks like that at seemingly random points. It did it less during first run parts with rapid override on. I finally traced it to the toolchanger operating. The head had a hydraulic counterbalance cylinder, and the toolchanger was hydraulic, when a TC solenoid valve opened or closed it changed the counterbalance pressure slightly and moved the head just enough to make a mark. When rapid override was on, the tool prestage movement was complete before it started to cut. Usually. So the only fix was to not prestage tools during crucial finishing cuts.
If I was visiting, I'd grab the head while it was cutting and see if I could shake it and make marks, just to look for movement as a start.
What kind of counterbalance does that machine have? Weights on chains? Do they operate smoothly? Anything disturbing them?
An air cylinder? Is it working right?

HAHA I like this post and I'll explain.
I was up on the ladder a month or so ago running some electrical lines for a lathe I moved last year when we got the fadal and never wired in.
when I was up there I noticed something odd about the chain on the counter balance. on link on the left side of the machine looked odd. I checked the gears and everything and everythign was smooth but one link was tracking different. it bother me for a day figured I would look into ordering a new chain. and I plum forgot all about it till you mentioned it above..

Damn MUD its 10:30pm and you have me thinking about driving down to the shop to see if thats a problem. I'm dead beat and just got home 2.5 hours ago. so whats going to happen is I'm going to jump in the shower hit the sack then sometime in the middle of the night wake up when the cats want outside and head down there. so much for a good nights sleep tonight LMAO.
And I just told my wife, remember when I saw that chain moving funny on that one link. Her reply was "Yeah what about that funny noise you heard thursday and yelled at me to stop running the machine you could her it but I couldnt"

Hmmmmmm

thanks for that idea.

 

[Herding Cats]

Not your issue but when tuning G08 and like on Fanuc machines this can become an issue.

This can also be caused by any backlash at all in the Z axis on a machine with some type of counter balance. The decel in a corner can be enough to change axial cutting forces which allow the tool to slightly move.

An interesting fact that I wasn't always aware of is that most higher end machines don't use a counter balance. The counter balance allows for use of a smaller screw, servo and drive but it also puts the Z axis close to an equalibrium where any slop will show itself. Having the weight of the Z axis can help to preload the system and make it more stable.

 

[lobust]

An interesting fact that I wasn't always aware of is that most higher end machines don't use a counter balance. The counter balance allows for use of a smaller screw, servo and drive but it also puts the Z axis close to an equalibrium where any slop will show itself. Having the weight of the Z axis can help to preload the system and make it more stable.

The primary driver for not having a counterbalance is speed. Some are worse than others, but all types of counterbalance put a hard limit on how fast the axis can move and how fast it can accelerate.

Curious if your Hitachi Seiki has one? It's a fast machine for it's age.

Only two machines in here have counterbalances on vertical axis. One old VMC (like early 90's old) with the typical chain/pulley/counterweight, and a largish bridge mill which has an active hydraulic counterbalance.

Both are slow on rapids, I think the Z is 15m/m on the bridge mill. Ram has a lot of weight in it though.

 

[Herding Cats]

The primary driver for not having a counterbalance is speed. Some are worse than others, but all types of counterbalance put a hard limit on how fast the axis can move and how fast it can accelerate.

Curious if your Hitachi Seiki has one? It's a fast machine for it's age.

Only two machines in here have counterbalances on vertical axis. One old VMC (like early 90's old) with the typical chain/pulley/counterweight, and a largish bridge mill which has an active hydraulic counterbalance.

Both are slow on rapids, I think the Z is 15m/m on the bridge mill. Ram has a lot of weight in it though.

No it does not have one. Rapid and feed while running G8 are 30m/m for a and 40m/m for xy.

Never thought of the acel deal but it makes perfect sense now that you mention it.

 

[Vancbiker]

Does this vintage of Fadal still use resolver feedback? If so I’d look at that too. I’ve seen some weird shit happen with flakey Fadal resolvers and couplers.

 

[Vancbiker]

Not applicable to the OP’s machine, but Mud’s post about hydraulic pressure causing trouble reminded me of a tough one I worked on long ago. A largish HBM would randomly bore off location by ~.0002-.0005” while line boring 777 main landing gear trunnions. Discovered that when the way lube pump turned on the oil pressure caused the head to move a tiny bit. Depending on how far out the quill was extended determined how much position error would happen in the bore. Ended up setting a spare m-code to turn off the pump cycle timer while line boring was being done. The complementary m-code reset it back on and a failsafe timer would turn it back on after 15 minutes in case someone forgot to program the reset m-code.

 

[Herding Cats]

Not applicable to the OP’s machine, but Mud’s post about hydraulic pressure causing trouble reminded me of a tough one I worked on long ago. A largish HBM would randomly bore off location by ~.0002-.0005” while line boring 777 main landing gear trunnions. Discovered that when the way lube pump turned on the oil pressure caused the head to move a tiny bit. Depending on how far out the quill was extended determined how much position error would happen in the bore. Ended up setting a spare m-code to turn off the pump cycle timer while line boring was being done. The complementary m-code reset it back on and a failsafe timer would turn it back on after 15 minutes in case someone forgot to program the reset m-code.

A tiny bit of hydraulic pressure is ridiculously powerful over a large surface area

 

[lobust]

Not applicable to the OP’s machine, but Mud’s post about hydraulic pressure causing trouble reminded me of a tough one I worked on long ago. A largish HBM would randomly bore off location by ~.0002-.0005” while line boring 777 main landing gear trunnions. Discovered that when the way lube pump turned on the oil pressure caused the head to move a tiny bit. Depending on how far out the quill was extended determined how much position error would happen in the bore. Ended up setting a spare m-code to turn off the pump cycle timer while line boring was being done. The complementary m-code reset it back on and a failsafe timer would turn it back on after 15 minutes in case someone forgot to program the reset m-code.

Interesting, my first VMC (a late '80's Hurco built on Japanese iron, if anyone can believe that Hurco and high end iron can exist in the same sentence) had an m-code for that same purpose and a page about it in the manual. It had really wide box ways and air over oil lube pump. It had all glass scales and was deadly accurate (positioning) all the time I had it.

I don't know if it was special ordered that way. It came from Ilmor Engineering who made racing bike and formula one engines.

 

[Delw]

Does this vintage of Fadal still use resolver feedback? If so I’d look at that too. I’ve seen some weird shit happen with flakey Fadal resolvers and couplers.

No this one doesnt. the 96 does, and yes I have seen some wierd stuff as well with resolvers and the couplers even wierder stuff.

 

[Delw]

Tech came in this Am, he checked for obvious stuff ie backlash on ball screws and thrust bearings. the z axis was only a tenth or 2 if that. they ya axis had about 1.5 x was under a .001
He did check the draw bar tension. on the problem Fadal it is 750 lbs, on the older fadal which I replaced over 2 years ago and thought it was going out again it read 1850, the hass was 1760.
we both dont believe its the draw bar as its only a .008 finish cut but it could be as the pressure is quite low.
I didnt have him check the chain as I was trying to get some 1st articles finished and my ladder is missing (think teh neighbors got it. so going to check that in a few mins.

 

[lobust]

Tech came in this Am, he checked for obvious stuff ie backlash on ball screws and thrust bearings. the z axis was only a tenth or 2 if that. they ya axis had about 1.5 x was under a .001
He did check the draw bar tension. on the problem Fadal it is 750 lbs, on the older fadal which I replaced over 2 years ago and thought it was going out again it read 1850, the hass was 1760.
we both dont believe its the draw bar as its only a .008 finish cut but it could be as the pressure is quite low.
I didnt have him check the chain as I was trying to get some 1st articles finished and my ladder is missing (think teh neighbors got it. so going to check that in a few mins.

If your drawbar force is that low, it's fair to assume that at least a couple of the bellevilles are broken, which could lead to them shifting under dynamic conditions.

I.e. it's possible (likely?) that your tool retention is unstable when the spindle is running.

 

[Herding Cats]

I will also add that the quality of tool holders as well as pull studs can make a big difference in drawbar force. If the taper is slightly undersize the pull force goes down, same for over length pull studs.

When testing drawbar force it is a good idea to install the brand pullstud you run into the tester. Once I played around trying different pullstuds (cheap vs quality) in the same machine and it caused the readings to change.

 

[Delw]

If your drawbar force is that low, it's fair to assume that at least a couple of the bellevilles are broken, which could lead to them shifting under dynamic conditions.

I.e. it's possible (likely?) that your tool retention is unstable when the spindle is running.
Oh most definately I am betting more than half are busted. If I remember right when I changed the older fadal I only had maybe 10 unbroken washers. only reason I noticed it was like cats said one tool holder wouldnt hold tight. it was actually loose in the spindle where you could shake it. I thought I had a bad tool holder but after checking them all I had 5 that did that and my finish was always decent too.

I will also add that the quality of tool holders as well as pull studs can make a big difference in drawbar force. If the taper is slightly undersize the pull force goes down, same for over length pull studs.

When testing drawbar force it is a good idea to install the brand pullstud you run into the tester. Once I played around trying different pullstuds (cheap vs quality) in the same machine and it caused the readings to change.

Yes it can.


I looked into the chain thing a little more this afternoon, I think its stretched alot as the right side sprocket and chain looks to be tighter than left side. soon as this job is finished I'm going to replace chain sprockets and the bearings.

I still didnt check level and I am sure thats off also.

 

[Delw]

I had a chance to disassemble the drawbar on the fadal today, thanks to a 20 min cycle time on the other machine.
tearing it down only took about 45 mins.
all the beville washers were in tack not one broken, they actually looked pretty new as well. There were 44 pcs in there (what is required for my set-up). The drawbar itself looked in really good shape as well.
One thing I forgot to do (because I assumed there was broken washers) was make sure they were positioned correctly when I was taking them out.
I didnt look in the spindle yet to see if the sleeve was damaged. so I am kinda curious why the very low pressure on the drawbar. as everything looks really good.
My other fadal had most of the washers broke so that was a easy one to find the problem.
I wonder if the washers could have just worn out? no gunk build up in there either.

Also with the cover off I could see the other part of the counter balance chain. something doesnt look right links are crooked so to speak. will dig into that more once I get the drawbar back together.
I'm going to have to buy me a force gage one of these days soon, even though I would rarely use it $200 bucks for someone to come out and check drawbar force will pay for one after 2-3 times.