Z Axis Sinking Mid Cut

Good Evening. Ive been having an issue lately where the Z axis begins to sink shortly after starting a facing pass. Ive read a lot of discussion about this issue and looked into many things suggested. the coupler of the motor and ball screw is good, the holding of the motor "while not running" is rock solid. the other night i was up on a step stool pressing with all my might and could not get it to budge while it was not moving. while it was moving however, it took considerably less pressure to get the motor to decouple and free spin downward. a slight lift up would recouple the motor and it would continue to move normally (except way off from where the machine thought it was now). i dont feel like this should be this easy to decouple the motor. if its rock solid while not moving, why is it so easy to decouple the motor when it is running ? possible frequency or steps/mm issue ? micro stepping set wrong ?

even while running the facing job where it typically sinks (without a bit and not engaging the work piece for safety) i could not get it to budge at all practically laying on the motor. even while the X and Y axis were going about their business normally.

but ... for some reason shortly after starting a facing cut, ive tried back and forth and i typically use a spiral cut, when it gets about half a bits diameter into the work piece, the Z axis begins to sink into the work piece. it does not do it when im using smaller bits, such as a .5mm tapered ball mill, or a 20' V bit. only when im running the 1/4" flat end mill does it do it. also, fresh end mills dont do it. after 3 or 4 facings, i can hear the tone of the cut changing and the cut gets rougher, and it starts to sink into the work piece.

its not the bit sliding in the collet, as after it happens if i rehome the machine and put it back at its Z depth the bit is still right where it should be. im "thinking" its something to do with the vibration of the slightly dulled bits causing the motor to slip, but i dont know of a way i could test this. or, perhaps the downforce of the duller bits is pulling the motor down into the piece, but i dont see this being the case if im practically laying on the motor when at rest and it doesnt move.

some specs:
-- highly modified genmitsu 3020 V1 running an upgraded Z axis and makita router motor.
-- using a recently acquired openbuilds controller and control software (which i absolutely love ! ordered another one for my second machine, dont think thats the problem as i was having this issue before upgrading to the new controller)
-- climb cutting with a 1/4" 0 flute carbide flat end mill at 30k rpm, at ~550 mm/min in pure copper 3mm depth (50%) and .3175mm stepover (5%). but ... over the past few months ive tried various depths and stepovers and it happens regularly even down to 1% stepover and 1mm stepdown. the only things i havent really played with is the rpm's and feed rate, but ive recently cut them both in half for when i recieve the new clamps i can test it out to see if it has better performance.

second question: as mentioned above, these 1/4" 0 flute carbide flat end mills "seem" to get dull within 3 to 5 facings, which leads to a significant drop in the tone of the cut, much rougher finish of the cut, and eventually leading to the drop in the Z axis ruining the part. i "think" this is related to the rather high rpm's and feed rate that im using ? if i cut it back to say 15k rpm's and ~250mm/min feed rate, will this help the bits last longer ? or even farther back to 10k RPM (lowest the makita router will go) and this low rpm i would have to go around 100 - 150mm/min feed rate. but im not sure if this low rpm would be adviseable with the makita router, since ive read that lower RPM's dont have as much torque and could get bogged down in the piece.

your inputs into this issue are greatly appreciated.

thank you

 

i can look into that for sure, but off the top of my head i dont think that would be the case as im using the openbuilds "meanwell" power supply, 24 volts and 14 amps. the openbuilds controller can go up to 4 amps per motor. the Z axis motor is a nema 23, while the x and y are nema 17's. i have the power set so that the motors are "warm" but not "hot". so i think the power levels are set right. and since im using the makita router as the spindle, it plugs into the main power on its own (through the IOT Switching relay, recently found this, absolutely love it. so much better than the watts based automatic power switch i was using before). it wouldnt be taking away from the power supply that feeds the stepper motors if it were a embedded spindle.

i appreciate your feedback though, ill measure the motors voltages while its running just to check that out

 

it is set to 255 yes. the workpiece is not very big so when it cooperates i can do the facing job in only a few minutes. but lately with this sagging issue ive been reducing the stepover mostly, so its taking ~20 minutes now. what would be a better tool to do the facing with ? keeping in mind my relatively small machine. max shank size is 1/4" and im doing metals such as copper being the hardest one, silver and "nordic gold".

thank you for your suggestion

 

simple 1mm depth spiral cut. although its not accurate for the problem im having since the 1k rpm only switches on the router which normally runs 30k rpm, the depth can vary from 3-6mm depth on some jobs, and ive tried a variety of stepovers from 1% to 50%.

 

ran the facing pass yesterday again with 15k RPM, 5% stepover and 250mm/min. apparently i was using a 2.5% stepover before. i also used a conventional cut this time rather than a climb cut. fresh bit, it cut ok. ill keep a close eye on the bit sharpness and try to rotate them out regularly, as well as keep playing with the rpm and feed rate. see if i can extend the life of the bits while keeping a reasonable cut time.

i was wondering if copper would "case harden" or work harden. because i cast these billets myself, i quench them in water after casting. i was wondering, assuming even that it would case harden them. using a climb cut, with too fast of RPM's i can see why that would tear up the bits. so using a more reasonable rpm with a conventional cut probably will increase the life of the bits a fair bit.

i appreciate your feedbacks. ill update if i keep having the Z axis drop issue.

 

thank you for your replies and the explanation of hardening. along with reducing the RPM's and feed rate, im also monitoring the sound of the cut with a spectrum analyzer to determine an exact point where it begins to cause a problem. as well as marking the bit to show how many times it has been used to keep track of sharpness / dullness. im an engineering type so i love detailed information like that hehe. ill keep playing with the speeds and feeds to extend the life of the bits and maintain a clean cut while just knowing that they wont last as long as i want them to lol. maybe getting in to bit sharpening will be a next step.