Milling Cast Iron

[dertsap]

after a better look at the code and at the picture i'd say its missed steps , either that or your part has moved ,
try dialing up your part dead nuts , try a few cuts then redial the part and see if it has shifted again

[M250cnc]

Well i think i will upset a few people with this statement.

Backlash compensation works in theory but not in the real world cutting difficult materials.

Why do you think quality builders put so much effort in to an engineering solution to backlash.

When conventional milling the axis motor power is being used to drive the cutter in the cutting process.

In climb milling the axis motor is being used to prevent the cutting axis being drawn into the workpiece

So when you are climb milling with the backlash hidden with compensation the cutter is being drawn into the workpiece unexpectedly this results in a sudden overload of cutting depth which results in you going off the toolpath appearing to be lost steps. This is probably why you broke a cutter.

So the answer is simple to fix your problem fix the mechanical backlash.

Phil

[Khalid]

Hi Curt,
Tomorrow i will test both methods (Climb and Conventional Milling) for a pocket ... I will also measure the Clamps from some reference with a dial...

Lets Suppose If I have Missed Step, then should the simulation results matches with the real part? You see i have the machining shifted but the results comply with simulation... The problem would be solved if i never stopped the machine and let it to finsih the part.. If the machined part matched what i required with a little shift then the phenomena would be much strange... I guess if i not stopped the machine, i would have deviation in the machined part...

I also think that the Axis which carries most of the load may have missed step.. In my Case Y-axis may be the culprit...but:
1- In CASE of CAST IRON Machining, I got deviation across Y-axis.
2- Incase of Aluminium Machining, My part is Shifted toward X-axis, which is the most Free axis i have.. Having very little load as compare to Y-axis..

I also ran the Y-axis Motor and pulled the Y-axis slide against the motor and i couldn't stopped that axis by hand...even i couldn'd slow it down. and i have no observation from the motor change of noise!!!...

For next test I put 200lb of weight on the top of X-axis slide and commanded Y-axis to move 4 inches and it went exact 4-inches+- 0.01... No Missed step...

Is there anything wrong with the shuttle acceleration or Backlash compensation?

[Khalid]

Well i think i will upset a few people with this statement.

Backlash compensation works in theory but not in the real world cutting difficult materials.

I am really a newbie in Backlash compensation.. I have a router and doing woodworking for the last 5-years with it and never used backlash compensation...
In my view, Backlash compensation only works at the time of change of direction of motor rotation...A lot of peoples here using backlash compensation with their machines and getting perfect results including Bobwarfield...

Why do you think quality builders put so much effort in to an engineering solution to backlash.

Because they sell their machines for good.. and they provide the quality product to the users that don't know how to compensate for the backlash time and again ..and they get Handsome amount of money from their sellings...
I know nothing so kindly pardon me in this...

When conventional milling the axis motor power is being used to drive the cutter in the cutting process.

In climb milling the axis motor is being used to prevent the cutting axis being drawn into the workpiece

So when you are climb milling with the backlash hidden with compensation the cutter is being drawn into the workpiece unexpectedly this results in a sudden overload of cutting depth which results in you going off the toolpath appearing to be lost steps. This is probably why you broke a cutter.

So the answer is simple to fix your problem fix the mechanical backlash.

Phil

Thanks for the informations... i learn daily something new from you guys.. From above statement i understand that i should use conventional milling?..Am i right...

[dertsap]

i wouldn't worry about how much weight your table can carry , but I'd be interested in how much weight it can pull
i agree the sooner or later you should deal with getting rid of the backlash because it will make your life easier , but in the meantime you'll have to learn how to work around it , I know how clever you are so with some experimenting with cuts you'll figure it out .
you need to eliminate all of the possibilities of failure so scribe a line at the head and the column and make sure the heads not moving , if you calculate you chip load you may be taking a heavier cut than what your setup can handle so try knocking your chip load in half and do some testing as to how much you can push your tools ,
your aware of the size of motors that I've got setup on my router and I dragged my boy half way across the table when I asked him if he thought he could hold the gantry back , but when I take too heavy of cuts in aluminum then I'll run into the same types of trouble . the last was doing 3d profiling on ally and found a problem with the way the y axis was mounted (whole new problem ) , point is that that you need to eliminate all possibilities for disaster

[Khalid]

if you calculate you chip load you may be taking a heavier cut than what your setup can handle so try knocking your chip load in half and do some testing as to how much you can push your tools ,

The same i think the stepper motor stall or loose steps.. The attached picture shows another PEN plotted with the same machine..The Gcode provided by RICH at machsupport forum..

The moves or IJK and started for different Z-moves.. I amazed to see how start and finish point coincide with each other.. Again i am in 0.01mm range..

I have to increase the voltages to my stepper from 36Volts to 48Volts... My stepper can withstand Voltages upto 80Volts DC with 4.2Amp/Phase.. Currently I am supplying 3.8Amp with 36Volts.. will it increase its strength by increasing the voltage?

[M250cnc]

Because they sell their machines for good.. and they provide the quality product to the users that don't know how to compensate for the backlash time and again ..and they get Handsome amount of money from their sellings...
I know nothing so kindly pardon me in this...

There are lots of users on here who built their machines with zero backlash "Me Included OK 0.03mm" because that is the best way when you want to machine difficult materials.

It is not a case of knowing how to use backlash compensation, that may work on soft materials or cutting AIR

Try running you gcode with no tool in the collet centre on the part as normal check at the end that you are back on zero

You will then know if the code is right

If the code stays on zero then you will know that the machine is at fault

Phil

[Khalid]

In above case no-load condition, The machine started/stopped at Zero.. and the results are within 0.02mm range...

So , In my point of view i am just loosing steps at high load on Y-axis..and thats my finding

[M250cnc]

In above case no-load condition, The machine started/stopped at Zero.. and the results are within 0.02mm range...

So , In my point of view i am just loosing steps at high load on Y-axis..and thats my finding

Backlashs for Axes:
1- X-axis: 0.005708 inch = 0.145mm
Y-axis: 0.013385 inch = 0.334mm
Z-axis: 0.0535433 inch = 1.36mm
Backlash speed of max%: 50%
shuttle Acceleration:0.08
Mill : climb mill

Its no coincidence that the Y axis has the biggest amount of backlash, which is where you have the greatest error.

There is no software substitute for a well engineered machine. That machine you have will work really well with no backlash.

Phil

[Khalid]

Hi Phil
Actually my Y-axis is the bottom slide which carries X-axis slide as well as the load of Work piece and also the cutting forces too.. Also I am using Box type slides that having maximum friction in it...This is the reason my Y-axis motor has more possibility of lost step..

As soon my Spindle is ready, I will show you the perfect results with the same machine with the same backlash compensation

Bottom Line..I am 102.999% sure my Y-axis motor is loosing steps

[neilw20]

First you must make sure stepper motor tuning is OK.
Without this you can often loose steps.

Make acceleration to quite slow.
Make program to go backwards - forwards - many times.
Make the maximum speed so that the axis will not stay in step and then reduce speed by 20%.
Then increase the acceleration until steps are lost.
Reduce acceleration by 35%.

Now for the backlash. Best to get rid of it.
But there is a simple fix, albeit ugly. Then you can climb mill too.

Put a weight and a pulley dragging table against the backlash. One for each axis.
There must be enough weight to more than compensate for the cutting loads.

After setting up the weights, check the motor tuning to ensure no lost steps.

Do you have limit switches? I hope so.
Do you reference the machine? You must always know where zero is.
Your impeller looks like missed steps, or slipping drive to screw.

Make some parts out of hardwood until it all works properly.
Crawl before you walk. It saves on cutters.

If the Z axis is counter balanced, and has a lot of backlash, then the cutter can pull the head down.
Either use too much, or not enough counterbalance, but don't let the head 'float' between it's backlash points.

Cutters with inserts don't have much top rake, if any, and this is best for cast iron.

Never use depth of cut more than half the diameter, in metal until you get really good at it.

Always ramp down. It is the dish angle on the cutter face that limits the ramping angle. The trailing edge of the cutter never cuts well.
Because you have such high backlash, very slow z feed when plunging helps. I use 3 flute cutters because they cant chatter by reacting against the opposite tooth.

Once the flexibility of the machine exceeds the chip load you will get chatter or unstable cutting. Flexibility includes backlash!

Best of luck. Find some good ball screws. Stiffness is EVERY THING.

[Khalid]

Thank You very much for your wisdom and help... Following is my feedback to your comments:

First you must make sure stepper motor tuning is OK.
Without this you can often loose steps.

Make acceleration to quite slow.
Make program to go backwards - forwards - many times.
Make the maximum speed so that the axis will not stay in step and then reduce speed by 20%.
Then increase the acceleration until steps are lost.
Reduce acceleration by 35%.
.

As the previous page shows the impeller drawn with the help of PEN... I measured the drawing with the CAD and comparision shows 0.03mm difference.. So My Steppers at no load condition and backlash compensation ON shows the best possible results with the given system..

My acceleration settings are 5 in/sec^2 for all the axes and the Velocity is 10IPM for all axis.
I will work on your suggestions tomorrow and will show you the results.. I have a little confusion...
My steppers are of 4.2A/Phase and can take up 80Volts.. Currently I am running these steppers with 3.5Amp and 36Volts?... My controller card can bear upto 60Volts DC... What do you think by increasing the voltages to 48Volts, should the performance of motors may increase?.. What if i increase the Voltage ?... Do Increasing the Voltages increase the Torque or WHAT???

Now for the backlash. Best to get rid of it.
But there is a simple fix, albeit ugly. Then you can climb mill too.

Put a weight and a pulley dragging table against the backlash. One for each axis.
There must be enough weight to more than compensate for the cutting loads.

After setting up the weights, check the motor tuning to ensure no lost steps.
.

I replaced the ACME threaded rods with used Ball screws... When turnining/twisting the ball nut with hand i didn't find any backlash... The ballscrews one end is supported with two thrust bearings and two radial bearings.. but I don't know how the backlash comes in the system...
I can't getrid off this backlash currently i am having

Hmm..You have given me good solution for the Backlash compensation but the extra weights on both slides will further limit my torque However the idea is very plausable and workable...

Do you have limit switches? I hope so.
Do you reference the machine? You must always know where zero is.
Your impeller looks like missed steps, or slipping drive to screw.
.

I have no limit switches , no Emergency Stop for now... However in future i am going to install ....
No I not reference the Machine... My reference is on the Actual Part where i set my X/Y/Z=0
You are 100% right the impeller is missed step example... At 7IPM i am loosing steps with beefy steppers (1200Oz-in Bipolar but I am using in Unipolar so getting almost 750Oz-in torque)...I wonder why... Again should increasing the Voltage will increase the Torque or WHAT?

WHAT IS SLIPPING DRIVE TO SCREW?..

Make some parts out of hardwood until it all works properly.
Crawl before you walk. It saves on cutters.

If the Z axis is counter balanced, and has a lot of backlash, then the cutter can pull the head down.
Either use too much, or not enough counterbalance, but don't let the head 'float' between it's backlash points.
.

Hmmm..Very nice point about the Z-axis counter weight... If i remove the counterweight then what would happen?... hmmm... I think i must try it tomorrow Thanks for such a nice advice..

Cutters with inserts don't have much top rake, if any, and this is best for cast iron.

Never use depth of cut more than half the diameter, in metal until you get really good at it.

Always ramp down. It is the dish angle on the cutter face that limits the ramping angle. The trailing edge of the cutter never cuts well.
Because you have such high backlash, very slow z feed when plunging helps. I use 3 flute cutters because they cant chatter by reacting against the opposite tooth.

Once the flexibility of the machine exceeds the chip load you will get chatter or unstable cutting. Flexibility includes backlash!

Best of luck. Find some good ball screws. Stiffness is EVERY THING.

I used Ramping to dig into the pocket... I agree with you i must RAMP very slowly , currently my RAMP settings are 2IPM... I am agree with the whole paragraph and will try my best to show you some more results...

Thanks NEIL for helping me

[neilw20]

Just drawing something doesn't mean you won't lose steps.
It just shows you didn'tlose steps.

The acceleration is very important, and you must run the tests at a faster and faster speed to FIND the limits.

You are probably too close to one of the limits.
You will almost certainly find each axis needs different settings.

The X axis only moves the table but the Y axis moves the total weight of X and Y, hence it has different inertia.

Running the supply voltage closer to the ratings of the drivers will increase your maximum speed, but the current limit settings control how much load the stepper can take before it slips.
How fast do you want to cut air? Do you really need the extra speed?

[Khalid]

Hi Neil,
I can run 15 IPM Rapid only... Upto 20IPM i loose steps... I somewhere studied that increasing the AMPerage means you increase the Acceleration and increasing the voltages mean increasing the speed...

So lets suppose if i increase the voltage upto 48V, and my RAPID increase to 30IPM.. (I will increase the voltage tomorrow and will show the actual results)... Will i loose steps at 15IPM?

[Khalid]

I drew with PEN just to check my steps/unit are correct or not.. I also want to check how backlash compensation works... I know the machine capabilities only shows when we cut the actual parts and not the Air alone....

One more thing i am thinking now... I want to remove some areas of My Y-axis box slides..so that it make less contacts with bottomm.. May be this help in loosing some frictional forcess...

[M250cnc]

Khalid

Based on your recent posts about top speed of 20IPM rapid there is something seriously wrong with your setup.

You are also mistaken on amperage and voltage, more voltage = more top speed, more amps = more power "Torque".

48 v isn't too bad but more is better.

The faster you set the acceleration the more chance for lost steps

You may have the jibs set too tight and you need plenty of lubrication on the slides, on my mill conversion i built a one shot lube system into the ways.

Jibs need to be adjusted first WITH THE MOTORS DISCONNECTED ALONG WITH THE SCREWS. So you can push and pull the table on the ways until you have no slop but no binding YOU CAN FEEL THE FIT

Then put back the way screws/ballscrews, turning the pulley by hand to check for smooth operation no binding.

Then refit the motors

On my machine "RF31" i got 4000mm rapid when i had missed steps "Motor Squeal" i cut that in half to run the machine at 2000mm rapid, the motors are 4nm at 70v 6amp so not that beefy.

Phil

[Khalid]

Khalid

Based on your recent posts about top speed of 20IPM rapid there is something seriously wrong with your setup.........You may have the jibs set too tight and you need plenty of lubrication on the slides, on my mill conversion i built a one shot lube system into the ways.


Jibs need to be adjusted first WITH THE MOTORS DISCONNECTED ALONG WITH THE SCREWS. So you can push and pull the table on the ways until you have no slop but no binding YOU CAN FEEL THE FIT

Then put back the way screws/ballscrews, turning the pulley by hand to check for smooth operation no binding.

Then refit the motors

Okay..very nice point about the Gibs tighteness... We have jib tightness adjuster wedges... I will check their tightness.. I don't know how much loose i should keep them so that their effects do not come in the part dimensions...

As suggested by you i will remove the motors+screws and loose the wedges and perform motions and tight the jibs accordingly so that there is no binding no slope.... May be the Jibs are too tight....

I MANUALLY THROW THE OIL ON THE SLIDES VISIBLE TO ME... I HAVE NO ARRANGMENT TO PUT THE OIL INSIDE THE CAVITIES PROVIDED IN THE GIBS... DO YOU PUT THE OIL INSIDE THE GIBS SLIDES?....
I HAVE TO MAKE SOME FLEXIBLE SYSTEM FOR OILING INSIDE THE GIBS... MAY BE THIS NO-OILING IS THE PROBLEM?????

On my machine "RF31" i got 4000mm rapid when i had missed steps "Motor Squeal" i cut that in half to run the machine at 2000mm rapid, the motors are 4nm at 70v 6amp so not that beefy.

Phil

WOW.. i sure want to see the picture of your machine and also the one-shot oiling system.

Thank you Phil for being with me and helping me... Your suggestions are very valuable to me...

[dertsap]

My steppers are of 4.2A/Phase and can take up 80Volts.. Currently I am running these steppers with 3.5Amp

you'll need to push more amperage to get those motors running the way that they should be , right now they are underpowered and your not getting the full amount of your necessary torque

[M250cnc]

Okay..very nice point about the Gibs tighteness... We have jib tightness adjuster wedges... I will check their tightness.. I don't know how much loose i should keep them so that their effects do not come in the part dimensions...

The accuracy would be greater with a bit of play and no lost steps than too tight and lots of missed steps.

I MANUALLY THROW THE OIL ON THE SLIDES VISIBLE TO ME... I HAVE NO ARRANGMENT TO PUT THE OIL INSIDE THE CAVITIES PROVIDED IN THE GIBS... DO YOU PUT THE OIL INSIDE THE GIBS SLIDES?....
I HAVE TO MAKE SOME FLEXIBLE SYSTEM FOR OILING INSIDE THE GIBS... MAY BE THIS NO-OILING IS THE PROBLEM?????

A one shot lube system is standard on all good machine tools some even have an auto pumped system.

Cheap/hobby machines are prone to high wear rates, lack way wipers. So anything that you can do to alleviate this is to be desired. The parts of the system are not that expensive, but are time consuming to install and that is the reason that they will cost a lot ready installed on machine tools.

WOW.. i sure want to see the picture of your machine and also the one-shot oiling system.

Unfortunately i would have to take my machine apart to show you and i am not doing that. But there are some build threads on doing a Oneshot system is is easy, but as i say time consuming, just do a search.

But a pull/lift of a plunger and all the ways are lubricated all at once ALL ALONG THE WAYS. You cannot do this manually.

I also built a table enclosure so that all cutting chips and coolant stay off the motors and the ways then all the way oil stays on the ways

This does make life difficult to work sometimes but the positives outweigh the negatives.

Phil

[Khalid]

Hi Phil,
I have checked the Gib Tighteness... Following were my observations:

1- The Y-axis Ball screw has 3inch dia pully installed.. I removed the belt and rotated the pully with One thumb and 3-Fingers... I was feeling the force of moving the Slide... The Screw has some spring action inside.. The wheel is turned first upto a limit with the screw and then screw nut takes the load and further after quarter turn the slide was moving...

I then adjusted the Taper wedge.. and now my Y-slide is moved with only two fingers and not enough force... For comparasion i would say i have reduced 50% force for movind the slide by loosening the taper wedge... I also checked the slope and it was un-noticable... The slide is very very Free now

2- The same procedure was adopted for my X_slide and it is also improved 50% from the previous...


I will calibrate the whole system again and will cut the part with it.. I will also check what maximum Rapid i can' get now without loosing steps...

I am thankful to you for helping me...