Slowly growing error, always in the same direction:
- In this case you have a part size that is slowly (a few tenths to a few thousandths per cycle) changing on a very regular basis. You can make a correction to the offset, then the machine will make a part that is close, then slowly drift again.
- Bad axis motor/encoder assembly: If the axis encoder is defective the slide can accumulate an error on every cycle. As the day progresses it is possible to minimize the error with offset changes. During the day you will not have much of a problem other than constantly having to change offsets. However when you start the machine the next day and home it, there could be a major offset needed. As you "correct" the error with offset changes the home of the tool is effected. When the slide is homed for the first time, the marker pulse re finds the original home location. While the axis encoder was misreporting slide movement offsets change the tool location.
- Loose motor/ballscrew coupling: Check the coupling for being loose or cracked. The coupling could be slipping on either the screw or motor shaft. We have also seen hairline cracks cause very intermittant problems. Normally when the slide rapids.
- Bad G code program: With the different versions of code out in the field there is always a chance that the code you have written could be causing a problem. Even if the code works in one machine does not mean that it will work in all. Sometimes the code could run a while and then mess up, you never know. A quick way to determine this is to eMail or fax a copy to a factory applications or service tech for assistance. We know what to look for and we would be happy to review your program and give you some feedback.
- Attachments loose on the bed of the lathes (retrofits only): Attachments mounted on the bed of a Hardinge have been known to move. Sometimes the gibs used to hold the attachment to the bed have become loose, cracked, or clamped on dirt and crud. Check it out.
- Bar feed pressure to high or to low: If the bar feed pressure is to high, the bar will walk its way out of the collet, if the pressure is to low the bar won't feed out enough on each cycle. In either case there will be variations in part length
Gross errors, almost always in the same direction
- In this case there are times the machine will hold size for a while, then make a large change in a sudden chunk.
- Broken or loose motor/ballscrew coupling: Check the coupling for being loose or cracked. The coupling could be slipping on either the screw or motor shaft. We have also seen hairline cracks cause very intermitant problems. Normally when the slide rapids.
- Attachments loose on the bed of the lathes (retrofits only):Attachments mounted on the bed of a Hardinge have been known to move. Sometimes the gibs used to hold the attachment to the bed have become loose, cracked, or clamped on dirt and crud. Check it out.
Random errors in a small range (up to +/- .01").
- Here you will see errors that are always within a few tenths to a few thousandths of the mean dimension. As you correct the offset, sometimes you are adding, sometimes subtracting.
- Cutting tool height of center: Height of center is a very common issue for random size changes. If you are using a 5/8" round shank holder, make sure the insert is on center. Do not rely on the flat on the top of the holder to bring the insert to center. As you tighten the screws holding the bar in place, rock the tool. This way the holder will come to center as the screw locks down. Also at issue is that the tool holder is correct as well as the bar itself. Take nothing for granted.
- Bad thrust bearing on ballscrew: The thrust bearing is what holds the ballscrew in place. If there is any play in the assembly you will see it as mispositioning error. Any error in the thrust bearing will be seen as backlash in the axis. Perform a push/pull test.
- Loose mechanical element, i.e. linear way, ballscrew nut: This can also be found with the push/pull test. Linear ways are bolted to be base and cross slide assembly, sometimes they come loose and need to be retightened.
- Broken motor/ballscrew coupling: If you find there is mechanical play in the system and it is not the thrust bearing or loose linear way, it could be the ball screw. This is rare but not un heard of.
- Axis motor mechanical problems: We have seen very rarely an axis motor that has come apart. The screws holding the motor together come loose. If this happens you can grab the axis motor and twist it some.
- Work holding for the part being cut: If the part being machined is not be held problems happen. Look for parts slipping in the collet or chuck. Also check part roundness or conformity to work holding, collet closer issues, bar wobble due to no spindle liner, Blank size - to big or to small in a collet creates problems.
- Tool holder issues: If there is a problem with the tool holder and it is not held firmly, vibration is a problem. If you are using OmniTurn tool holders look for a small hairline crack or dirt under the holder. Check the insert pocket, make sure it is not cracked or damaged.
- Bad X axis slide brake: There is a spring loaded brake on the X axis of GT-75's and GT-Jr's. Air pressure is used to pull the brake away from the screw. When the servos are turned off the spring loaded plunger is allowed to rest on the coupling. If the solenoid or piston are defective there can be drag on the X axis. This could cause mispositioning. To test this theory, remove the brake from the X axis. Check with tech support on how to do this.
- Bad cutting tools: A defective or incorrect cutting tool can cause size changes. If the material is not being cut it can cause too much tool pressure. The increased pressure will push the part and slide around.
- Problems with blank material: Sometimes the material itself varies. We have seen material with hard spots or variations cause major changes in size without reason. Cutting a few parts in a known, easy to cut material will confirm or eliminate this issue.
- Blank sizes vary: If the blank has variations in size or shape (i.e. bent) it can cause variations in the finished parts.
- Amplifier out of balance, check loop gain: If the amplifier for the axes are off there could be a variation in part size. The loop gain in particular should be checked. If the loop gain is to loose the slide will be "mushy" and could vary every cycle. Tightening up on the loop gain will make the slide stiffer. If the loop gain is to tight the slide will vibrate creating bad finish or oscillation.
- Rapiding to close to the part, hitting it before machining: If you rapid to close to the material before changing over to feed rates it is possible to hit the part. This could cause the part to move in the work holding, bend the part, damage the cutting tool. Make sure the tool stops at least .03" away from the rough blank material.
- Chip buildup around cutting area: Chips wrapping around the tool or work piece can create a who series of problems. One of them is to cause coolant starvation at the cutting point. Change the cutting cycle to brake up the chip or change its characteristics.
Large errors when you start each morning
- If the machine tends to hold size OK during the day, then when you restart the machine the next morning it is off by quite a bit.
- Bad motor/encoder assembly
- Bad encoder counts during the cycle: If you are offsetting a tool all the time to maintain a dimention, the first time you run a program after the slide has just been homed there could be a problem. Run the CTRL-H/CTRL-C test.
- Bad home switch - the slide homes where ever you start the homing sequence. This means the marker pulse for finding slide home is not working. You can see this if you move the slide away from the 0 mark on the scale. Normally when you home the slide it will drift back to the mark. With a bad home switch it will not move, it makes home where ever it happens to be.
- Bad G code program: With the different versions of code out in the field there is always a chance that the code you have written could be causing a problem. Even if the code works in one machine does not mean that it will work in all. Sometimes the code could run a while and then mess up, you never know. A quick way to determine this is to E-mail or fax a copy to a factory applications or service tech for assistance. We know what to look for and we would be happy to review your program and give you some feedback.