Grinding Spindle

[derek hall 1Participant @derekhall1]

Hi,

I am currently building my Quorn and have reached the stage of making the grinding spindle assembly.

However I was wondering if anyone has built Mr D Broadley's grinding spindle (available as a kit from Hemingway Tools) and wondered if this would be an easier and possibly better option?

Kind regards

Derek

[derek hall 1Participant @derekhall1]

[John PParticipant @johnp77052]

Haven't built the Broadley spindle but looking at the pdf
drawing here " Broadley F.pdf – Home Model Engine Machinist Forum"
I would stick with the original Quorn designed spindle.
The Broadley designed unit is similar to the Quorn design
probably a little easier to make i think the spring box arrangement
of the Quorn is better than the dished washer set up of the Broadly
design.
I built my Quorn spindle to the Chaddock design in the book
35 + years ago never had it apart, still runs quiet and smooth.

There are many ways to make a better spindle that is stiffer that
the Quorn ,these 3 here are slightly larger than the Quorn use only
standard SKF eco bearings 20 lb preload run up to 31,000 rpm
at 65 deg c .

John

[derek hall 1Participant @derekhall1]

Hi John

Thanks for reply, looks like the Quorn spindle is the way to go.

Kind regards

Derek

[John PurdyParticipant @johnpurdy78347]

John

I'm curious as to why you think the spring box arrangement for preloading the bearings is superior to the disk springs washers. Could you elaborate?

When I built mine I used the disk springs as I felt it would be easier and produce a more uniform preload. In use it has been perfectly satisfactory, and I feel was easier to make.

Derek

I have a PDF of the drawing of my modified spindle using the disk springs also a short write up of its construction. If you are interested I can send them to you.

John

[duncan webster 1Participant @duncanwebster1]

In the distant past I worked on design of gearboxes for small gas turbines. They used wavy washers and angular contact bearings. Shaft speed was ~80,000 rpm, so first gear shaft speed would be 16-20,000 rpm.

One potential problem with the spring box arrangement is getting equal load on all the springs

[John PurdyParticipant @johnpurdy78347]

Duncan

That was my concern. When I got my castings from MES, included were 6 springs for the spring box. A cursory inspection of them showed that there was a difference in the spring rate between them. This would mean that if made as drawn, the preload would not be uniform around the bearing. . Whether this uneven preload would make any difference in the operation of the spindle in the type of service that the Quorn spindle is used for I don't know, but that is why I went with the disk spring design. In the case of the turbine gearbox you mention I suspect that uneven loading on the bearings at those speeds and prolonged running would very likely cause premature bearing failure.

John

[John PParticipant @johnp77052]

Posted by John Purdy 14/11/2021 18:20:18
John

I'm curious as to why you think the spring box arrangement for preloading
the bearings is superior to the disk springs washers. Could you elaborate?

When I built mine I used the disk springs as I felt it would be easier and
produce a more uniform preload. In use it has been perfectly satisfactory,
and I feel was easier to make.

———————————————————————–

Hi John,

I don't think there is any problem with the disk springs being used
for the pre-load on these bearings it is more the way they have been
used directly against the rear bearing ,in this way there is little support
for the bearing, in the D.H. Chaddock design and also most spindle
bearing designs use a sliding sleeve which is sprung loaded against
the bearing outer ring since the sleeve is longer than the bearing
diameter it ensures the outer ring is held square within the spindle body.

The photo of spindles shown earlier are similar to the Chaddock
designed spindles except for all the machining operations are
done at a single setting with no turnaround setting up including all of the
labyrinth seal parts with the exception of the rear end cap retaining
threads in the main spindle body. The spring box length is about 2 x the
bearing diameter ,each of the 3 springs the pressure is set individually
by drilling the holes too deep and dropping in a spacer to set up the
spring pressure .

In the Quorn Universal tool and cutter grinder book Prof Chaddock
makes reference to this very point about the spring box on page 53
in which he writes,

" The other bearing is firmly clamped on the inner ring but the outer ring
is pre-loaded by means of a spring-box .This enables the spring
pressure to be carefully controlled and providing the spring box
has a length approaching its diameter and is a close sliding fit in its
housing ,ensures that the spring pressure is applied squarely to the
outer ring"

The washers as in the Broadley design which has recently been
seen on the Quorn user group website albeit printed in reverse
has none of these features and is entirely reliant on the springs
to ensure this squareness.I suppose another failing with this
arrangement , a spring box is in close contact with the end of the
bearing ring and as such allows a path for some of the heat from
the bearing ring to travel away ,with the cone spring arrangement
there is only a a very small contact area all of the heat has only the
contact of the bearing ring on the outer sleeve.

The OP had asked for an opinion on the the 2 designs of spindle
for his machine. This is perhaps a more detailed response to that
original question.
In any event i suppose he will have to decide this for himself.

John

[John PurdyParticipant @johnpurdy78347]

John

Thanks for your explanation. Your reasoning makes good sense and I can see why it might be superior in that aspect. Perhaps a better way to utilize the disk springs would be to have a sliding sleeve 2 x as long as the bearing diameter bearing on the bearing outer race with the other end bearing against the disk springs to provide the preload. This would alleviate your concern of little support for the outer race and provide for heat transfer, at the same time providing a uniform preload around the circumference of the outer bearing ring. Just a thought.

John

[John PParticipant @johnp77052]

Hi John,

I had come to that same conclusion using the disk springs in the way that you describe ,the Quorn spindle is not that demanding , it is lightly loaded and does not run that fast to be a problem ,the 3 spindles seen earlier run to 31,000 rpm they are only limited by the drive belt which gives up due to the centrifugal force flexing the belt as seen here ,the belt has so much tension it needs some effort to fit over the pulleys but turns into chewing gum at high speed.

John

[John PurdyParticipant @johnpurdy78347]

Just as an aside, ( and not implying anything ) the Unimat SL lathe spindle uses two A13 magneto bearings ( the same as the Quorn ), and the preload is provided by a pair of disk springs ( >< )bearing directly on the outer race of the bearing at the pulley end. Out of curiosity I was going to measure the preload that is used on the Unimat but my spring scale that I used when I did my Quorn spindle seems to have grown legs and disappeared!

John

[Mark RandParticipant @markrand96270]

Posted by John P on 14/11/2021 22:27:02:they are only limited by the drive belt which gives up due to the centrifugal force flexing the belt as seen here ,the belt has so much tension it needs some effort to fit over the pulleys but turns into chewing gum at high speed.

For future designs, would it be better to use nylon reinforced flat belts? They are nowhere near as stretchy as the round polyurethane belts.

[John PParticipant @johnp77052]

Posted by Mark Rand 15/11/2021 10:33:49

——————

Posted by John P on 14/11/2021 22:27:02:they are only limited by the drive belt
which gives up due to the centrifugal force flexing the belt as seen here ,the belt
has so much tension it needs some effort to fit over the pulleys but turns into
chewing gum at high speed.
——————
For future designs, would it be better to use nylon reinforced flat belts? They are
nowhere near as stretchy as the round polyurethane belt.
——————————————————-
Hi Mark,

Most probably .

Like most things some compromises have to be made,
manufactured belts are fixed sizes and as such what they fit on has
to be designed to make them fit ,that would have caused me somewhat of a
problem as to accommodate the swing up head seen here no adjustment
is possible to tension the belt.

As the spindle has a working speed from about 4,000 to the
31,000 rpm accomplished via belt changes and inverter
regulation additional problems with other fixed belt sizes would undoubtedly
be present, the heat joint belting gets around these complications just by
being able to be made to suit, the inconvenience of the belt flexing
only occurs above the 31,000 rpm which is more than enough for
the internal grinding for which it is designed.
Looking on the lathes web site at some of the past type grinding machines
this machine does rather well in spindle speed terms some examples:-

Crystal lake grinder internal spindle speed 17,500 rpm
Jones and Shipman Universal Grinding spindle 11,600 to 16,200 rpm
Jung internal grinding machine 30,000 rpm
Mipsa Swiss 23,000 rpm
Duplex toolpost spindle 18,000 rpm
Jones and Shipman 36"x 12" 3080 to 33,000 rpm

These are just a few ,i am sure there will be some more ,i guess there is
an upper limit on driving a spindle by belts,it would seem that i am already
up at the top end compared with some of these well made names.
I have an air spindle that runs at around 50,000 rpm for those times
when that this may be needed.Since the machine was originally designed
as a cylindrical grinder and is also a surface grinder along with the
internal facility some inevitable compromises have been made ,it just
what you have to do to build something like this.

John

[Author]

Posts