Pratt Burnerd 4 jaw Chuck jaw alignment

I've recently acquired an 8 inch Pratt Burnerd 4 jaw independent chuck. It is stamped 'Made in England' and I think it was manufactured in the 80s. It seems to be unused or virtually unused as there are no marks on the chuck whatsoever and the D1-3 camlock fitting shows no signs of ever having been attached. The jaws have a perfect ground finish and are unmarked with no detectable play in their slots. When I fitted the chuck to my lathe to test trueness I found the chuck body to run true to <0.01 runout.both on the periphery and on the face. The clamping surface of the jaws however are not parallel. They are the opposite of 'bell- mouthed' as in the pic below. The taper is consistent on all four jaws. This may not matter too much for many applications but is definitely apparent when trying to grip a long and accurately machined workpiece.

What is the explanation for this? Were the jaws deliberately given a taper to accommodate wear or do I just have a badly made chuck? I am suspecting the latter.

I do reaslise that this taper would be easy to correct by the in situ grinding of the chuck jaws and I may well resort to this.

I think I would strip and clean the chuck and relubricate, also check the register is running true and remachine if necessary. I would eliminate all possibilities before attacking a lightly used chuck with a regrind.

Mike

I wonder if they were deliberately tapered to give a better fit when they are tight. You can test this theory, by turning a length of aluminium for a parallel smooth finish, and then clamping it nice and tight in the chuck. Do the jaws up slowly and as evenly as possible, and then look at the jaw marks on the bar to see how even they are. All chucks and jaws are subject to a little flexure when tightened.

I would prefer this to bellmouthing any day.

I've already stripped, cleaned and re-oiled. Apart from some congealed grease there was nothing to remove, not even a speck of swarf. The register is an integral camlock D1-3 and the chuck body runs perfectly true. All looks good apart from the issue described...

No way the jaws are supposed to have a 0.1 taper along the gripping face as your sketch describes. Something is amiss.

With that much gap you might be able to check the jaws with an engineer's try square to see if the gripping surface is square to the base of the jaw.

If you have a surface plate, you could set each jaw up on a known good angle plate, or a cylindrical square and run a dial indicator mounted on a base over the horizontal surface to check squareness between the gripping surface and the guiding surface.

If they turn out not square, regrinding is the only cure. If they do turn out square, then the guiding surfaces in the chuck are machined at an angle and still the only cure I can see is to regrind the jaws.

I think it's doubtful that it's the jaws themselves at fault. Is this one of those lightweight 4-jaw chucks with a shallow section and pockets cast in the back? They bend quite easily if you go daft with the chuck key. I would sweep the chuck face and then the jaw guides with a dial gauge mag-mounted to the cross slide. If the face of the chuck is dished then you can grind the jaws but they might not then be straight at another diameter.

I don't know if I believe this myself but .... is there any chance this was a "special" produced by Burnerd for a particular customer for a particular purpose and then not used?

Is there any sign that the jaws have, in fact already been reground? It's just that the fact that all four jaws are the same and to the same extent seems significant to me.

The chuck body runs true on all faces to <0.01mm runout with no distortion of the body. There is nothing to suggest the jaws have been reground (but who knows).

I've removed the chuck jaws and held them with the clamping face against a true vertical surface as below. Running a dti along the slot in the jaw confirms that the slot is not at right angles to the jaw face. All four jaws are out by the exact same ammount. This seems very odd. That degree of consistent inaccuracy cannnot be manufacturing error.

.

I have no knowledge of the design specification, so cannot comment on whether they are [all] right or wrong ... but I would beg to differ with your assertion ^^^

A good machine, wrongly set, can produce a consistent error.

MichaelG.

If the jaws have not already been ground, if laid on a flat surface, would the outer ‘lands’ not be found to be parallel? Not so much distance over which to measure, but should show a difference?

Sounds like a manufacturing fault to me. If so, probably explains why the chuck is still in unused condition!

Even the best manufacturers have occasional quality problems due to faulty machines or human error. Inspectors don't catch everything especially if they operate a random sampling system and it's far from unknown for factory rejects to be retrieved from a skip and sold privately. Not everyone in 1980's Britain was honest!

But is it the jaws or the chuck itself that's faulty? If the jaws are at exact right angles, maybe the chuck's T-slots are off.

Dave

Edited By SillyOldDuffer on 21/10/2019 10:43:49

If as you say the chuck is otherwise pristine, although they have no obligation, my next port of call would be Pratt B for some friendly "help".

Oh yes it can. Even with "finest" British engineering. Back in 1971 a whole batch of Triumph 650 twin cylinder motorcycles were unleashed on an unsuspecting public that kept on scrubbing out one side of both bores. Turned out, after much gnashing of teeth, that whole batches of cylinders had been bored out of square to the cylinder base surface. Machine was set up wrong (Old Cyril who usually ran it retired and never told the lads how to jam the wooden wedge in the wornout slideway to set it square?) and away they went. Mass produced errors of perfect consistency.

Your fix is new jaws or regrind the old ones. Might be worth contacting PB but doubtful they would do much on something older.

Four thou (.1mm) is not a lot to grind. A dremel held in the toolpost would do the job while the jaws are clamped on a cloverleaf plate to spread them in the operatiing position.

Edited By Hopper on 21/10/2019 12:10:58

Edited By Hopper on 21/10/2019 12:12:10

Many thanks for all the suggestions. Sounds like I just have a 'Friday afternoon' chuck. I'm told most days were Fridays in some factories in the 70s and 80s !

Will try regrinding the jaws.

It's a FOUR jaw independent chuck. Not necessary.

To the OP, have you also checked the other (stepped) side of the jaws as you might need to use this one day?

I'm pretty sure that nobody's checked the jaws of a four jaw independent before. Plenty of three jaw scroll chucks get tested. I would hold back from any grinding until other jaws have been tested properly.

When three jaw scroll chucks are ground it is important to preload the jaws first, which removes any backlash and tilts the jaws in their slots. Have the jaws in question been preloaded in the chuck body before testing, I think not.

Has a freshly ground three jaw chuck jaw been tested for squareness, I very much doubt it.

Edited By old mart on 21/10/2019 16:00:40

Number of jaws makes no difference. You still need to preload them to hold them in the working position and to stop them moving about during grinding.

The OP says the taper on his jaws causes problems when he tries to grip long jobs so I think it's safe to say there is a problem with the jaws.

.

... You just need to find that four-leafed clover

MichaelG.

I have I think 3 sets of 4-jaw chuck jaws of unknown provenance. Put up some basic dims of your jaws and I'll check the ones I have. In the unlikely event that one of those set match, you can have it.