gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
080921-1558 EST
Last Wednesday I received a call from a process engineer about an electrical failure on a pinion preload adjust machine. This is a machine that automatically runs the nut onto a pinion on an axle assembly until the correct bearing preload is achieved. Actually what is measured is the pinion bearing drag torque via a 30 # force transducer at a fixed radius from the pinion centerline and it is assumed that this is a fairly accurate estimate of the axial force on the bearings. At correct drag torque, 20 #-in, the loadcell force is about 7 #.
This particular machine was manufactured in 1978 and probably began production that year. There is a large gear box on the machine that rotates the pinion at one of two speeds, 90 or 30 RPM, and simultaneously rotates the pinion nut at one of three speeds relative to the pinion and controlled by two electromechanical clutches. The net result is that the pinion nut can rotate at about 6, 2, 1, 1/3, 0 RPM relative to the pinion.
The clutches are a standard product modified for this application. These are each about 80 # and worth maybe $2000 each. The work to change clutches is about 3 men, different trades, for 4 to 8 hours. Therefore, you do not change a clutch unless there is good evidence of the need.
This gear box has been on the machine for the last 30 years. The two speed drive motor has never been replaced. However, the clutches will have been changed a number of times because of internal spring failures. The clutch type and its modification has been used in this kind of application since 1973. In that time I do not believe there has ever been a failure of a clutch coil. A machine cycle time is about 24 seconds. On average during this machine cycle the clutches probably cycle 2 to 3 times. At 120 parts per hour and maybe 4000 hours per year we have about 500,000 parts per year. For 30 years maybe 15,000,000 parts have run thru the machine.
The clutches are somewhat over 100 ohms and are powered from a full wave bridge circuit whose input is 120 V and the clutch coil has an RC snubber circuit in parallel. P&B KUP type relays are between the bridge rectifier and the clutch coil. On breaking the circuit the coil voltage rises to a peak of maybe 1000 V.
This is background for the troubleshooting exercise.
The call I received from the process engineer indicated that this machine's fast clutch would blow the clutch power supply fuse when the fast clutch turned on. This fuse is a 1.5 A AGC (standard blow) and never blows unless there is a short in the wiring or two clutches come on at once. Note: the information I received by phone is at least secondhand. Things are usually different when you get to the job.
The plant electrician had tried various tests and could not find any shorts and the clutch coil resistance was normal.
None of this information points to a particular cause. At this point I doubted that the clutch was the problem. This of course was remote by telephone. Secondhand information may be quite different than what is available from a firsthand source.
These days you do not go on a service call without a purchase order. Finally got a PO late Thursday. On Friday I went to the plant.
I had an electrician, the one who had been working on the machine, a job setter, and a machine repairman to work with me.
The clutch power supply in this station has never been replaced in the 30 years. However, the plug in relays have been replaced many times, probably 2 to 3 times per year. The bridge rectifier follows the fuse and is not the problem because it is powered all the time. All diodes were good. We rechecked the clutch coil resistances and these were 105 or 106 ohms and steady.
Next put a clamp-on DC current probe around the lead to the fast clutch wire. Energized the relay and the fuse did not blow and the current was in the ballpark of the expected value.
Next we disconnected the clutch coil from the power supply. This machine does not have a manual means to rotate the drive motor. The ohmmeter was connected to the leads to the clutch, and we used clip leads to energize the motor contactor. The first thing I noticed was squeaking from the gear box. This is not good, likely a bad bearing. With the spindle rotating we noticed a fluctuation of the clutch resistance. On the electrician's Fluke we could see changes of 85 to 127 ohms. By the way, when talking with the electrician he indicated that the fuse did not always blow immediately when the fast clutch was energized. This was critical information I did not get over the phone and thus I had not believed the clutch was the problem.
So now we know there is a problem only when the motor is rotating. The next step was to disconnect the clutch leads at the last point before the clutch. Connected the ohmmeter to these leads on top of the machine. Reran the test with rotation. Same varying result. Thus, problem is in the gear box.
Next the oiler drains the oil from the gear box. The machine repairman then has to remove the front cover plate. After removal he felt the residual oil and noted there was a lot of grit. Immediately this was an explanation for the squeak.
Without rotation there is no variation in the coil resistance reading, it is at 106 ohms. Then the electrician by hand moves the clutch plates and the resistance varies. Clearly the problem is in the clutch assembly. I looked at the clutch in a hard to see location and there were metal chips clinging to the outside of the clutch. Note: there is considerable residual magnetization of the clutch parts. We can expect that these metal chips have gotten inside the clutch assembly.
The end result. Clearly this whole gear box has to be rebuilt, and new clutches are required. Once there are metal chips throughout the assembly there is no easy and reliable solution except to replacing everything.
My guess is the coil in the clutch has not been damaged, but that chips have gotten into the leads going into the clutch. This we could not see because they are in a conduit. But this may in fact mean that the coil is damaged because these leads are potted into the coil.
My initial conclusion by phone that the clutch was not the problem was based on prior history and that the implication was that the fuse immediately blew when the clutch relay closed.
See my website http://beta-aa.com/pa_plot.html for a plot of nut torque and preload torque for an adjustment cycle.
.
Last Wednesday I received a call from a process engineer about an electrical failure on a pinion preload adjust machine. This is a machine that automatically runs the nut onto a pinion on an axle assembly until the correct bearing preload is achieved. Actually what is measured is the pinion bearing drag torque via a 30 # force transducer at a fixed radius from the pinion centerline and it is assumed that this is a fairly accurate estimate of the axial force on the bearings. At correct drag torque, 20 #-in, the loadcell force is about 7 #.
This particular machine was manufactured in 1978 and probably began production that year. There is a large gear box on the machine that rotates the pinion at one of two speeds, 90 or 30 RPM, and simultaneously rotates the pinion nut at one of three speeds relative to the pinion and controlled by two electromechanical clutches. The net result is that the pinion nut can rotate at about 6, 2, 1, 1/3, 0 RPM relative to the pinion.
The clutches are a standard product modified for this application. These are each about 80 # and worth maybe $2000 each. The work to change clutches is about 3 men, different trades, for 4 to 8 hours. Therefore, you do not change a clutch unless there is good evidence of the need.
This gear box has been on the machine for the last 30 years. The two speed drive motor has never been replaced. However, the clutches will have been changed a number of times because of internal spring failures. The clutch type and its modification has been used in this kind of application since 1973. In that time I do not believe there has ever been a failure of a clutch coil. A machine cycle time is about 24 seconds. On average during this machine cycle the clutches probably cycle 2 to 3 times. At 120 parts per hour and maybe 4000 hours per year we have about 500,000 parts per year. For 30 years maybe 15,000,000 parts have run thru the machine.
The clutches are somewhat over 100 ohms and are powered from a full wave bridge circuit whose input is 120 V and the clutch coil has an RC snubber circuit in parallel. P&B KUP type relays are between the bridge rectifier and the clutch coil. On breaking the circuit the coil voltage rises to a peak of maybe 1000 V.
This is background for the troubleshooting exercise.
The call I received from the process engineer indicated that this machine's fast clutch would blow the clutch power supply fuse when the fast clutch turned on. This fuse is a 1.5 A AGC (standard blow) and never blows unless there is a short in the wiring or two clutches come on at once. Note: the information I received by phone is at least secondhand. Things are usually different when you get to the job.
The plant electrician had tried various tests and could not find any shorts and the clutch coil resistance was normal.
None of this information points to a particular cause. At this point I doubted that the clutch was the problem. This of course was remote by telephone. Secondhand information may be quite different than what is available from a firsthand source.
These days you do not go on a service call without a purchase order. Finally got a PO late Thursday. On Friday I went to the plant.
I had an electrician, the one who had been working on the machine, a job setter, and a machine repairman to work with me.
The clutch power supply in this station has never been replaced in the 30 years. However, the plug in relays have been replaced many times, probably 2 to 3 times per year. The bridge rectifier follows the fuse and is not the problem because it is powered all the time. All diodes were good. We rechecked the clutch coil resistances and these were 105 or 106 ohms and steady.
Next put a clamp-on DC current probe around the lead to the fast clutch wire. Energized the relay and the fuse did not blow and the current was in the ballpark of the expected value.
Next we disconnected the clutch coil from the power supply. This machine does not have a manual means to rotate the drive motor. The ohmmeter was connected to the leads to the clutch, and we used clip leads to energize the motor contactor. The first thing I noticed was squeaking from the gear box. This is not good, likely a bad bearing. With the spindle rotating we noticed a fluctuation of the clutch resistance. On the electrician's Fluke we could see changes of 85 to 127 ohms. By the way, when talking with the electrician he indicated that the fuse did not always blow immediately when the fast clutch was energized. This was critical information I did not get over the phone and thus I had not believed the clutch was the problem.
So now we know there is a problem only when the motor is rotating. The next step was to disconnect the clutch leads at the last point before the clutch. Connected the ohmmeter to these leads on top of the machine. Reran the test with rotation. Same varying result. Thus, problem is in the gear box.
Next the oiler drains the oil from the gear box. The machine repairman then has to remove the front cover plate. After removal he felt the residual oil and noted there was a lot of grit. Immediately this was an explanation for the squeak.
Without rotation there is no variation in the coil resistance reading, it is at 106 ohms. Then the electrician by hand moves the clutch plates and the resistance varies. Clearly the problem is in the clutch assembly. I looked at the clutch in a hard to see location and there were metal chips clinging to the outside of the clutch. Note: there is considerable residual magnetization of the clutch parts. We can expect that these metal chips have gotten inside the clutch assembly.
The end result. Clearly this whole gear box has to be rebuilt, and new clutches are required. Once there are metal chips throughout the assembly there is no easy and reliable solution except to replacing everything.
My guess is the coil in the clutch has not been damaged, but that chips have gotten into the leads going into the clutch. This we could not see because they are in a conduit. But this may in fact mean that the coil is damaged because these leads are potted into the coil.
My initial conclusion by phone that the clutch was not the problem was based on prior history and that the implication was that the fuse immediately blew when the clutch relay closed.
See my website http://beta-aa.com/pa_plot.html for a plot of nut torque and preload torque for an adjustment cycle.
.
