Three Jaw Chuck Runout Problem
I am very new to the machining and this forum, I am only 15 years old. I bought a lightly used 3 jaw chuck off of PM. As expected it has quite a bit of runout because the backplate was not made on my lathe. When a piece of T6 Al tube is chucked up I measured 0.01" of runout about 1/4" away from the jaws. When I got it I took the chuck apart and cleaned it out and lubed it so there shouldn't be any foreign material in the scroll.
Here is a drawing of the run out as I measured it on the back plate:
Another problem is that how the holes in the backplate were drilled it is extremely difficult to bolt it to the chuck, it requires a lot of fudging and messing around to get it on. I have tried all the different combinations of holes in the chuck and the backplate and none of them are any better.
What do I need to do in order to get it to run true? I have already spent a while searching the archives of this forum and have not found anything that really fits my situation.
Beautifully worded and imaged question, it's a delight to (try to) answer when you have the facts laid out. Welcome to the forum, you'll do extra well, because you communicate so well.
You're going to have to re-machine the wobbly face and the register (presumably but not necessarily the one you show as 0.003" TIR)
There are several ways (lots actually, as this thread will no doubt reveal) to recover the lost fit.
One is to bypass Go, do not collect $200, and go direct to a home-made "Tru-Set" or "Set-Tru" or "Grip-Tru" deal, using grubscrews through the chuck body periphery. If you search on these terms (with and without hyphens) on this forum, you'll get somewhere
Edit: one more thing, you can't really use an extrusion to check runout, although for large amounts it will give you an indication. Ideally it wants to be something moderately hard with a ground finish. But lacking that, you can simply turn the outboard end of a piece of bar (or heavy wall tube) accurately round, and then turn it around in the chuck and grip and indicate on the turned portion.
If you have a surface grinder at your disposal - you can pull the jaw that is high (reads low on indicator runout test) and grind the inside chucking surface down a bit.
H Ross in '08.
Now more then ever!
The first thing to do is to see if the runout of the chuck back plate is repeating every time you remove and install the back plate. If there is variation, then look for an error in the fit of the back plate on the spindle threads.
If the runout is consistent, then turn the back plate true on the face and register for the chuck. Remove the least possible amount of material. You can also turn the plate OD to help with balance at high speed.
Now the chuck will be slightly loose on the back plate register. Attach the chuck and leave the mounting screws slightly snug. Chuck up a good true bar of appropriate diameter, like a steel dowel or a new chucking reamer or endmill. Extruded aluminum tube is not a good teast bar. Pay close attention to whether all the jaws close evenly on the bar. Sometimes the jaws just grip at the rear end instead of along the whole bar. Then the bar will wobble, even if the gripped part is running perfectly true. See if you can get the outside of the chuck body to have zero runout by lightly tapping the chuck with a wood block to move it on the back plate. Once the chuck body is running true, snug the screws some more. Now check the runout on the test bar. Then try a different diameter test bar and see how the runout compares. If the bar runout is consistent for different diameters, you can try bumping the chuck again to make the bars run true, and let the chuck body run out. At that point, there is not much you can do to improve things except to grind the jaws. PBA will grind jaws and overhaul chucks, but it might be better and less expensive to get a new chuck and back plate.
Thank you all for your quick responses.
I do not have access to any machine tools other than my lathe so for the moment surface grinding is out of the question.
I will try your suggestions tomorrow after school and post the results.
If you haven't the facilities to tackle this quite yet, you could get a hold of some thin shim stock, (preferably steel not brass but brass would do OK) and turn the register very carefully until you can just persuade the chuck to fit with three short lengths of shim equispaced. This is fiddly to get them to stay in place while you offer it up- thinner shim is easier. You can roll it to a curve with a steel dowel or similar, and flare it a bit on the entry edge, then stick it in place with a small dab of sticky grease
fit chuck to backplate
When you say the holes in backplate are difficult to bolt to chuck it makes me think that maybe who ever drilled holes might not have got them drilled true to the tapped holes and
they could be binding a little . Way to fix is to open size of holes up a little so chuck can move a little on back plate. If you are going to make a set true chuck you have to open the
mounting holes so it can move some. I would open holes about .01 and recheck it then.
This is not a terrible amount of runout, I have seen chucks worse than this. Have you checked to make sure that they cleaned it before they put it on the machine? That could cause the problem.
Nice post. When I was your age I worked in my Dad's manufacturing plant grinding flashing off castings and running a drill press with a tapping head and single point threading on lathe.... Oh, wait. You were asking about a chuck.
I think your approach, checking the back plate for runout is correct and should advance you to the point where you can get good runout measurements on chucked items.
Even so, L Vanice is correct when he says that aluminum is not really the best test bar. Once you get the back plate trued, if you've a piece of drill rod, an unused large drill bit, a chucking reamer, you can use the smooth portion of those as test bars.
But also check the jaws. One thing that makes chucks not really repeatable is when the grooves and ridges (in the chuck casting, and the jaw, respectively) get worn so that the end of the groove is bigger than the middle ("bell mouthing"), and the jaw runs a bit loose in the groove when you chuck something small. This can make repeatable chucking near impossible.
If, when the chuck jaws are almost closed, you can wiggle the jaw axially (towards the headstock and tailstock) more than about 0.001, this may be a problem. In bigger chucks, it's worth it to hard chrome plate and regrind grooves parallel and to size. In smaller ones, the view aint worth the climb.
If the backing plate is trued, you have the chuck mounted right, and you have a tight fit of the jaws in the grooves, the last step is to ensure that the tips of the jaws themselves aren't themselves bellmouthed. IF so, you can fix this problem using grinding, and this operation has been covered elsewhere in this forum. The symptom would be that a chucked bar can be wiggled fairly easily, or that a chucked piece indicates out of round, when everything else on the chuck tests concentric.
I have a six jaw six inch Buck, and the body grooves and both inner and outer jaw lands, AND the jaw tips are fried. Probably my only really bad ebay purchase.
Fix the undersized mounting holes FIRST. They will throw your mounting
scheme all to heck. The screws should have *generous* clearance. If you
have to shoehorn them in forget getting anything right.
Step 1 Fix the mounig holes first they should be lose to.01 clearance.
Step 2 Mount the backplate to the spindle.
Step 3 Machine the backplate most all chuck back plates are MENT to be machined thats the only way to insure it will turn dead true .
Step 4 Mount the chuck not tight.
Step 5 Put a test bar in the chuck do not use a piece of alum tubing its not round you want to use a ground piece of steel if you have one .
Step 6 Now move the chuck untill your test bar runs true.This is the reason you want the mounting bolts lose so the chuck will move a little.
Step 7 Tighten the mounting bolts done.
Ryland, you did a very good post and gave all the necessary info.
As I understand, you have .010" runout with the chuck on the lathe and the runout of the mounting plate is .003" side to side on the mounting step so the chuck jaws have an issue seperate from the mounting plate.
What you need to do is see how much slop is in the threads when you mount the plate on your spindle. If the threads are a loose fit you will never get a repeatable runout. With loose threads and the runout of the steps on the plate if you do any machining to repair the plate your wasting your time. To turn the steps under size with loose threads is faulty thinking. If the threads are tight on the spindle then you could machine the steps and then shim it or just bolt it to the chuck after opening the bolt holes. If you do this and the chuck shifts under load it will then be out even farther than it is now. You could dowel pin the mounting plate and chuck after you have it dial indicated in and that may keep it from shifting. I think two dowel pins would work.
Your only real way to correct this is to make a new mounting plate. It don't have to be cast iron. You can use cold roll steel. There are several threads on this site about making a new chuck mounting plate.
Is there enough meat on the step to allow four set screws to bear on it radially? Could the the chuck body be drilled and tapped radially for adjustmet that way? If that could be done, you could fabricate an adjustable function.
The real reason I'm posting here is that of all the posts above, I did not see a specific reference to examining the adapter plate threads very carefully. Before you do any machining, find a dentist's inspection mirror and a flashlight. Use them it to inspect the internal threads for embedded burrs. A small piece of hard steel will stick itself to the internal threads of a cast iron backing plate and stay there for years. WWQ
Bad chuck plate
Ryland, Welcome aboard. There has been a lot of good advice on your chuck problem. Let me give you one more solution. The mounting holes should be freed up or relocated if possible. With the back plate secure on your lathe, turn the boss down about .010 on a side. Leave rough tool marks on surface and DO NOT contaminate. Mix up a batch of JB weld and apply to boss thick enough to finish machining to correct size. I usually like size to size for a tap fit. This material is OK for a compression fit (7,000 to 8,000 PSI) and not too good for a running fit. After assembly recheck "0". Usually within .003 is as good as it gets unless you make it a "Buck" chuck. Good luck. Let us know how you made out. Dave.
Another way to go at this is to just face off that locating shoulder completely and mount a disc of whatever fairly thick sheetmetal you have on hand where it used to be. Mount with three or four screws and turn the disc down for your new shoulder.
If you re feeling froggy, just get a hunk of steel or iron slightly bigger than the chuck and make a new backplate. Only real challenge will be threading the blank. Main thing to remember is that, using the three jaw, you need to LEAVE THE BLANK CHUCKED UP AT ALL TIMES. That means when you need to test your fit on the spindle, you will leave the work chucked, unscrew the whole works from the spindle and turn it around for a test fit.
The rest of it is no harder than fixing what is wrong right now.
I have opened up all the mounting holes and machined the backplate down so now all the faces have 0.001" or less runout. Now I have to bump it into position. Is a section of 1-1/2" cold rolled 12L14 steel round ok for a test rod?
No, cold rolled round bar is not perfectly round and straight. Use a hardened and ground cylindrical tool, like a chucking reamer or long endmill or tap, or a steel dowel or drill blank. Drill rod is centerless ground, but is soft, so it might be bent or slightly out of round. It all depends on how accurate you need your measurements.
A piece of TGP or drill rod would be better. Most CR is not very accurate. Don't forget that you have some run out in the chuck jaws.
Try indicating the chuck body in and then checking how far out the rod in the chuck is out. You will have to decide if you want to indicate the chuck body to "0" or the rod in the jaws to "0". If you indicate the rod in the jaws you may be better off.
Also, try to get a 1/2" rod, 1" rod, 2" rod and 3" rod and they only need to be 4 or 5 inches long. Start with the 1/2" rod and indicate it in by bumping the chuck untill it's as close as you can get and tighten the bolts. Then put each next size larger in the chuck and see how much the run out changes if it does at all. If all the different size shafts run +/-.001 then your jaws and scroll are in good shape.
I enlarged the holes until they all lined up no problem, then I turned all the surfaces of the backplate to 0.001" of runout or less and tried mounting the chuck and tapping it to the center but I just couldn't get it. So I decided to try the JB Weld idea that Dave Gischel suggested. I cast a ring of JB Weld around the shoulder and turned it down until it just barley fit into the chuck with no play. The holes lined up really nice. I used a tap held in the jaw to measure the runout, I used the 1/2" or so of the shaft that extended beyond the jaws and got 0.006" of runout, I also chucked up the a piece of 1.5" CR and got the same runout. 0.006" is almost half of what I was getting before so it is defiantly an improvement. Do you think the remaining runout is in the jaws?
The fact that the runout is the same at different diameters narrows it down to either a difference in the jaws (maybe one jaw) or the register of the chuck is out of true (unlikely).
The scroll (the flat spiral plate which drives the backs of the jaws) is probably not your problem - if it was, you would expect different runout at different diameters. However, it may pay to check the "phase" of the runout: mark the outside of the chuck to indicate whereabouts you are reading the highest point on the test piece, for each diameter.
If the phase angle is different (ie the marks are well apart on the chuck) at the two diameters, the fact that the runout is the same value is a red herring, and a coincidence, and cannot be remedied by grinding one jaw or remachining the register in the back of the chuck.
If the phase is the same, you could do the latter remedy by chucking a truly straight and round bar (if your 3 jaw chuck is 4", 1" bar would be OK) in a 4 jaw chuck and get it perfectly true, then turn the 3 jaw backwards and grip it onto the bar so you can skim the register bore true.
Then you could scrape off the JB weld and do the same again.
Before putting this into action, though, it would pay to check the amount and phase of the runout for a couple more diameters.
I wouldn't normally recommend this remedy but it's more likely to be within your grasp than (say) grinding one or more jaws.
I also wouldn't do this to any other than a utility grade chuck, and certainly not for use at over, say, 1000 rpm.