R8 Insert Deflection

[shooter123456]

There are parts of a machine that need to be precise. There are parts that don't need to be precise. This test indicates the upper bore of a machine doesn't have to be precise.

RJSakowski posted in another thread that manufacturers are in fact not precise with this dimension:

Two things: Bridgeport designed the R8 taper, so their specifications would be THE specifications. The fact that there isn't a standardization is not an indication that the dimensions aren't critical.

Second, the lack of consistency between manufacturers is also not an indication that the dimension is not critical. There are differences in holding force, longevity, run out, surface finish, etc between different collets. The fact that some work better than others can likely be attributed to the fit between the collet and spindle.

And as someone else noted, the fact that manufacturers go through the trouble of grinding both those surfaces can't be ignored. If it really made no difference, they wouldn't be doing it.

 

[Bill Kirkley]

Shooter 123456, as stated, all of these machines are derived from the Bridgeport standard, so great care is taken in maintaining that standard.

What may not have been considered is weather or not some of the specs are critical. Machines evolve over time based on new findings through testing, innovations, materials, etc. I believe it is possible no one ever considered whether or not the dimension in question is critical.

The test clearly indicates a load on a cutting tool does not deflect or tilt the tool holder. If you accept that, then the ID of the upper part of the spindle is in fact not critical.

 

[matthewsx]

Shooter 123456, as stated, all of these machines are derived from the Bridgeport standard, so great care is taken in maintaining that standard.

What may not have been considered is weather or not some of the specs are critical. Machines evolve over time based on new findings through testing, innovations, materials, etc. I believe it is possible no one ever considered whether or not the dimension in question is critical.

The test clearly indicates a load on a cutting tool does not deflect or tilt the tool holder. If you accept that, then the ID of the upper part of the spindle is in fact not critical.

I seriously doubt that engineers at Bridgeport and Precision Matthews have never considered if any spindle dimension is critical, most likely they have specs and tolerances that parts are manufactured to in order to keep their products performing as designed.

Both companies are still in business so if you have any questions perhaps asking their customer service departments would give a definitive answer.

John

 

[RJSakowski]

There are parts of a machine that need to be precise. There are parts that don't need to be precise. This test indicates the upper bore of a machine doesn't have to be precise.

RJSakowski posted in another thread that manufacturers are in fact not precise with this dimension:

Manufacturers may be precise in this dimension not not accurate. Precision and accuracy are two different animals.

The problem that arises isthat given no "official" specifcation outside of Bridgeport, other manufacturers had to reverse engineer. the parts. This typically involves making a number of measurements of existing "good" parts and attempting to recreate the Bridgeport spec. If a large number of parts are measured, there is a fairly good probability that one can have a good value for the internal diameter of the top of the Bridgeport socket. There is also a small but finite probability that the value will be out in left field.

If a manufacturer wanted to determine a spec. for the Bridgeport R8 socket, he should measure a sufficiently large sample of Bridgeport machines. This is not really a practical situation and more then likely various manufacturers would skimp on sample size, increasing the probability of an erroneous value. Note that the measurements should only be made on Bridgeport machines as any other machine could be tainted by there own reverse engineering.

Measuring R8 collets/adapters is not the best way to set the standard as well as they should only be manufactured by Bridgeport or one of their subcontractors privy to the internal Bridgeport spec. and even then, the specification for fit with the socket is unknown.

 

[shooter123456]

Shooter 123456, as stated, all of these machines are derived from the Bridgeport standard, so great care is taken in maintaining that standard.

What may not have been considered is weather or not some of the specs are critical. Machines evolve over time based on new findings through testing, innovations, materials, etc. I believe it is possible no one ever considered whether or not the dimension in question is critical.

The test clearly indicates a load on a cutting tool does not deflect or tilt the tool holder. If you accept that, then the ID of the upper part of the spindle is in fact not critical.

That is based on the assumption that the only function of that face is to prevent static deflection of the tool. It could play a part in concentricity, repeatability, wear, vibration resistance, etc. It could also be to help transmit radial loads to the upper spindle bearings instead of just the lower ones. I could also be that it doesn't become important until more force is exerted.

I think what everyone is saying is that one test to measure one factor isn't enough to conclude that some of the largest machine tool manufacturers in the world have failed to consider that they could be saving time and money with no detriment to performance if they just made that part oversized and didn't worry about it.

 

[Bill Kirkley]

Unless someone can find fault with my methodology, the test shows that a load applied to a tool in an R8 tool holder, resulting in a torque of 946 inch pounds, causes no deflection of the threaded end of an R8 tool holder.

It appears there is disagreement as to the ramifications of this finding.

 

[mksj]

Clearly the specs for the R8 collet are well known and published by Bridgeport, Hardinge and no doubt multiple other manufactures of premium collets. The spec. for the R8 collet OD limit is 0.9495", the degree that a collet producer adheres to this spec, their tolerance limits, and their QC varies by manufacture. On Hardinge R8 collets the spec. is +0.0000 -0.0003. When I measured my Lyndex and Royal (Crawford) R8 collets they were all 0.9495 or slightly under. The spindle ID is manufacturer specific. If your collets meet this spec. and the spindle did not fit it, then clearly the spindle was faulty and should have been replaced under warranty. If you now feel that the spindle OD is bored concentrically and too the proper dimension then end of story. I have yet to see that you evaluated any axial deviation measurement at say 3-4" from the collet with an accurate test gauge. I can assure you that you need a gauge that can measure down to 0.0001" to have any meaningful results. Repeat the test pulling the drawbar up and down several times. If they are reproducible then you are should be OK as too the vertical collet alignment in the static mode.

Given that the R8 collets has been around for a long time requirements for manufacturing are quite tight, I seriously doubt the it is just haphazardly decided as to what the spindle bore should be and that the upper dimension can be whatever they decide on that day. If this was the case the surfaces would not be precisely ground to interface with the corresponding spindle interface. I have had no "zero" issues with quality collets and arbors AND have only had issues with generic collets like the Interstate for whatever reason. Other problems I have seen is the guide pin is set too deep (or is too wide) or the threaded portion of some arbors is not deep enough preventing the drawbar from not being able to pull the collet completely vertically.

If you look at other similar collets like the 5C, the rear ground surface is precisely ground despite being pulled back by a threaded portion similar to the R8, any small deviation in axial play results in measurable skew. The back interface on my Bison 5C chuck is very tight, previous Chinese chucks less so and it showed more inconsistency in axial deviation. There are also limitations to R8 system because of the nature of the collapsible interface of collets as compared to other collet systems/spindle interfaces which offer far better concentricity and rigidity.

The R8 drawbar is not precisely located and it is made to pull up on the collet, and not provide a lateral position of the collet. In my mill it rattled around until I modified the spacers. A small deviation at the top of the collet when the collet is drawn up of say 0.001" would result in a corresponding runout/wobble at the cutter tip. When milling there are vibration/oscillating forces, with corresponding movement/deflections at the end mill tip. The forces and effects can only be determined in a dynamic situation and is specific to the cutter type, collet, clearances, rigidity, etc. It is a known fact that runout of endmills corresponds to endmill life in a production environment, as well as the accuracy of the finished product. Clearly these are not issues in your situation.

 

[mikey]

Unless someone can find fault with my methodology, the test shows that a load applied to a tool in an R8 tool holder, resulting in a torque of 946 inch pounds, causes no deflection of the threaded end of an R8 tool holder.

It appears there is disagreement as to the ramifications of this finding.

Bill, just a few points to ponder:

• How exactly did you enlarge your upper spindle? I'm curious because it appears you did this without removing the spindle from the machine.
• The upper spindle is intended to resist radial forces. The force required to break an end mill varies with the size of the end mill, of course, but the force to break a 1/2" end mill is in the neighborhood of 116,000 - 200,000 pounds-force/inch squared. I've broken 1/2" end mills before so I know that the forces experienced by a cutter are considerably higher than what you're using in your tests. They can be even higher if you use a large face mill or even a large diameter fly cutter due to shock loads. You asked about ramifications ... here it is.

You seem to be trying to make yourself feel better about modifying your spindle. That's okay because as you said, its good enough for what you do and nobody can gainsay that. I think what at least some of us are saying is that another option is to just replace the spindle and use good quality collets ... and then move on, with a lesson having been learned. I assure you that we're trying to help you.

 

[RJSakowski]

Clearly the specs for the R8 collet are well known and published by Bridgeport, Hardinge and no doubt multiple other manufactures of premium collets. The spec. for the R8 collet OD limit is 0.9495", the degree that a collet producer adheres to this spec, their tolerance limits, and their QC varies by manufacture. On Hardinge R8 collets the spec. is +0.0000 -0.0003. When I measured my Lyndex and Royal (Crawford) R8 collets they were all 0.9495 or slightly under. The spindle ID is manufacturer specific. If your collets meet this spec. and the spindle did not fit it, then clearly the spindle was faulty and should have been replaced under warranty. If you now feel that the spindle OD is bored concentrically and too the proper dimension then end of story. I have yet to see that you evaluated any axial deviation measurement at say 3-4" from the collet with an accurate test gauge. I can assure you that you need a gauge that can measure down to 0.0001" to have any meaningful results. Repeat the test pulling the drawbar up and down several times. If they are reproducible then you are should be OK as too the vertical collet alignment in the static mode.
View attachment 303846

Given that the R8 collets has been around for a long time requirements for manufacturing are quite tight, I seriously doubt the it is just haphazardly decided as to what the spindle bore should be and that the upper dimension can be whatever they decide on that day. If this was the case the surfaces would not be precisely ground to interface with the corresponding spindle interface. I have had no "zero" issues with quality collets and arbors AND have only had issues with generic collets like the Interstate for whatever reason. Other problems I have seen is the guide pin is set too deep (or is too wide) or the threaded portion of some arbors is not deep enough preventing the drawbar from not being able to pull the collet completely vertically.

If you look at other similar collets like the 5C, the rear ground surface is precisely ground despite being pulled back by a threaded portion similar to the R8, any small deviation in axial play results in measurable skew. The back interface on my Bison 5C chuck is very tight, previous Chinese chucks less so and it showed more inconsistency in axial deviation. There are also limitations to R8 system because of the nature of the collapsible interface of collets as compared to other collet systems/spindle interfaces which offer far better concentricity and rigidity.

The R8 drawbar is not precisely located and it is made to pull up on the collet, and not provide a lateral position of the collet. In my mill it rattled around until I modified the spacers. A small deviation at the top of the collet when the collet is drawn up of say 0.001" would result in a corresponding runout/wobble at the cutter tip. When milling there are vibration/oscillating forces, with corresponding movement/deflections at the end mill tip. The forces and effects can only be determined in a dynamic situation and is specific to the cutter type, collet, clearances, rigidity, etc. It is a known fact that runout of endmills corresponds to endmill life in a production environment, as well as the accuracy of the finished product. Clearly these are not issues in your situation.

Interesting that the Hardinge drawing has the half angle of the taper when the angle of the taper is actually 16º 51' for a half angle of 8.425º.

From what I have been able to find, the diameter of the cylindrical section is .9495 +0000/-.0003" or .9496 +0000/-.0003". However, I have been unable to find any direct reference to an actual Bridgeport spec. After extensive searching, it appears that many others have tried and failed as well.

FWIW, my two machines have a socket dimension greater than .9494" (slip fit) . A .9495" cylinder is a light press fit. A 9505 was a no go. Neither are Bridgeports though.

 

[mksj]

Specific to RJ's comment about the Bridgeport spec. for the R8, there is information that was posted elsewhere which is informative/provides the reference spec. and also some discussion of the female receiver dimensions although that seems to be more opinion then a standard. I am using quotes below to provide credit to the individual who posted the information. This is a similar situation to many of the import 123 blocks, which although the external dimensions are accurate, the holes for bolting them to other blocks of fixtures are undersized and the bolts do not pass through these holes. Regardless of who made the the R8 collets, if they were no larger than 0.9495", then your spindle dimension may have been ground incorrectly. How this is addressed is clearly your choice.

R8 Taper socket dimensions? - RC Groups

R8 Taper socket dimensions? uk.rec.models.engineering

www.rcgroups.com

"Ok we need to look at a bit of history here to get the full picture.
The R8 collet was devised by Bridgeport back around WW2, before that they were using the B&S collets or sometimes the Morse. All this is rather hazy due to licensing agreements.

Because the R8 was classed as a proprietary taper it was never listed in places like Machinery Handbook, to my knowledge unless it's in V27 it still isn't.

Many people have listed various drawings for the R8 some very crude and some with a few major dimensions on them as a guide. many of the collet manufacturers have them on line.

The first recorded entry I can find on the FULL specs is in a 1952 edition of the National Machine Tool Builders manual. This was a professional body of interested people who went about setting various standards, machine noses and tapers being one.

Incidentally the initials NMTB is still prevalent today as a standard taper although it has also been superseded by the term INT

Here's a copy of that drawing hosted by Scott Logan of Logan lathe fame. http://www.loganact.com/tips/r8.htm

So now we had a standard and all was fine - until.............

Sometime with the last three of four years with the import of tools from the far east a certain American manufacturer who was building his version of a machine with an R8 taper sent a sample to China for collets to be made. Unfortunately due to a lack of published standards his spindle wasn't standard so the Chinese working with what they had made collets up to fit this spindle. The Yank was pleased as he now had a supply of collets but in the process had ****** up the whole shooting match.

So what has now happened is that in certain provinces they are making collets to the Twonk spec and in other provinces they are making collets to the correct spec.

The importers know this problem but it depends on where they are buying from whether they admit to knowing about this.

Over the last couple of years I have bought about 12 pieces of R8 tooling from collets to boring heads, all have fitted my Adcock and Shipley built Bridgy with no problems so watch where you buy from.

Regards,
John Stevenson
Nottingham, England."