My mentor had vast experience and was gracious enough to not only teach me
all methods, but wise enough to point out the pitfalls of using each one.
I have him to thank for what little knowledge I carry with me today.
It's been a very revealing experience to find that so many grind by methods
different form the one you and I prefer. I can say with total honesty that
grinding by my recommended procedure is the best way, hands down, assuming
finish is important. That would apply to even small, dry grinders.
Sorry to say I've never touched a double disk grinder, although I have seen
one. If by chance you manage to corner some information about one, I'd be
keenly interested in hearing what you discover. We had no need for one at
the missile facility, otherwise we'd have had one.
. Even with a mask you will
Heh! I can't even machine it without smelling it for what seems eternity.
For some reason, there's something in iron (gray and ductile) that doesn't
agree with me, or at least my sense of smell. I used to run a job that
had ductile iron dogs, which were a part of a drive I used to build for
Seagrave Corp.. They were used for opening large doors, or gymnasium
partitions. I ran a vacuum cleaner off the back side of the spindle and
discharged outside to keep from breathing the dust. Half of the operation
was boring the parts for a large bushing. The vacuum setup was a natural,
with virtually 100% of the dust ending up in the vacuum.
"Harold and Susan Vordos" wrote in
I've read acouple of posts where you gave out good advice on surface
grinding and offhand tool grinding. We had similar teachers.
The company I work for imported a surface grinder from Taiwan for a time.
We used to gring hardened 52100 steel to a mirror finish at the machine
tool shows for a demo. All you need is the right wheel, balance it, dress
it properly, and grind using the method you describe. Worked every time.
These weren't great machines either. They were better quality than I
expected but they weren't Mitsui's by any stretch.
Actually we sell one. We have had some issues with the wheel we are using
and the manufacturer hasn't been a lot of help. It's a CBN wheel and it
has been chipping out on the edges during cycle. The trouble seems to be
on parts where we are grinding single sided rather than double disk. I've
got some other projects where customers are wanting us to run a part and
I'm not comfortable buying another wheel from the current supplier at the
moment. I need to hold .0001" parallelism and .0001" flatness. Thickness
is wide open at .0003" total. I have to hold a 1.67 Cpk running these off
so there is no room for variation. When the wheel chips on the current
job the chip gets dragged across the wheel and my dimensions will jump
about .00006". sixty millionths may not seem like much but it's enough to
fail the run off. To say nothing about a very expensive wheel falling
apart. If you want to see the machine it's here:
it's not in English, and we haven't added it to our web
I've been away from machining since '83, so I'm painfully behind in what's
happening today, including knowing anything about a Mitsui, but a friend
that is working as a mold maker was talking about one of them recently.
They're apparently a super nice grinder. Your comments on procedure
should be very enlightening for those that don't subscribe to what I call
*proper* grinding, verifying that which I've been espousing.
Wow! Very impressive looking machine, and a total surprise. The one I'd
seen had horizontal spindles with a carrier for the part which oscillated
back and forth between the two wheels. Very old technology, to be sure.
One of the jobs the shop was running with is was the slide portion of a
I'm not familiar in the least with CBN wheels, considering they were not in
use when I retired from (commercial) machining. I gather that such a wheel
is more or less a necessity for the work at hand, but I'd be inclined to use
a different wheel, assuming that would be possible. You didn't mention the
material being ground.
Funny, talking about .000060" and having it be a concern. Yeah, it
doesn't sound like much to worry about, but when you have no tolerance to
speak of, it obviously spells the difference. I think it's hard for the
average machinist to grasp the concept of fine work. So much so that I've
commented more than once that it takes a special kind of guy to be
successful in the grinding department. Some guys never really make the
grade, yet may have years of shop experience.
About the chipping. I can't help but wonder if it's not being caused by
your parts chattering slightly. Your operation being a total stranger to
me, could it be that your coolant plays into the problem? For typical
grinding, the level of concentration of lubricant (in the coolant) isn't
critical, more for rust prevention than anything, but when you use machines
that rely on surfaces for support (like when grinding on one side only) it
could be you need more lubrication. We always ran a higher level of
concentration in the centerless grinder for that very reason. It's
interesting that it happens when grinding single sided, which prompts my
chatter question. My other suggestion would be feed rate, but without
understanding how the machine operates, I'm not sure I'd be in the ball
I looked at the lathes offered by your firm. Sure makes me breath easier
to know that if I was really in need that there are still reasonable
machines on the market. I was particularly impressed by the one model
that offered 2,500 RPM at the spindle. I'm running a Sag 12 Graziano that
I bought new in '67, and I"m hoping it will see me through to my dying days,
but I'd not hesitate to buy something similar (new) if the need arose. I'd
rather it didn't come from Harbor Freight, if you get my drift!
"Harold and Susan Vordos" wrote in
This machine is the smallest one they make. But you know how it is, the
smaller the part the smaller the tolerance. It's tough to see in the
photo but the fixtures are actually gears. So the parts rotate from the
inside of the wheel to the outside while rotating around the wheel.
eventually during the cycle all of the workpiece surface gets ground by
every bit of the wheel surface. The flatness and parallelism are really a
function of the bearings in the machine, and the machines ability to
resist deflection. The feedrate is determind by pressure, same as flat
lapping. The machine has auto sizing via an in process gage which is made
The one job we had trouble with we were single side grinding. The
material is 52100 low 60's Hrc. I can't remember exactly and don't have
the drawing here at home. CBN is the only way to do this type of
To tell you the truth it's surprisingly easy to achieve these tolerances
on this machine. (When the wheel doesn't fall apart) One of the other
parts we did double disk grinding, we ran for 24 hours and had a total
thickness deviation of 1.6 microns and flatness and parallelism never
exceeded 1 micron. I'm willing to bet that if we re-cleaned and rechecked
the parts that were at the extremes we would find that the deviation was
That's our current thought. The parts that the customer has supplied have
way too much variation, so the tall ones are under pressure and the short
ones might be jumping around as they pass beyond the edge of the wheel.
Either that or we have a bad wheel.
You might not be familiar with what we are doing but you're thinking is
dead on. The wheel manufacurer was certain it was the coolant. We were
using an equivelant to the one they sell. We then switched to their own
brand and the wheel still chips. They were claiming that the damage was
already done by the coolant we were using. Then I pointed out that the
lower wheel didn't chip and its flooded in the coolant. Hmmmm. Well then
you must have crashed the wheel. I'm telling you, that never happened. If
it did I would just get another wheel and get on with my life.
Like I said the feed rate is controlled by pressure. We are starting out
at low pressure until we get to a point where we are certain that the
wheel is contacting all of the work. Then it shifts to a higher pressure
and then back to low as it finishes. The lubricity idea is interesting.
I'm going to look into that some more. When we are single side grinding
we are using hardened fixtures that ride on a hardened plate. The part we
are grinding locates on a shoulder in the fixture. So there is
opportunity for something to stick. The speeds are fairly slow though and
the flood of coolant is immense. I am thinking that we could re make the
fixture so that the workpiece doesn't come out from under the wheel as
the fixture rotates around. The problem then becomes we would have to
desin a dummy part for the in process gage. (this part is very small) The
double disk ground workpieces use a dummy part mounted on the periphery
of each fixture for gaging. These dummy parts are larger in diameter than
the part being ground.
Thanks. I don't have anything to do with the "manual" side of the
business, but I can get knee mills and lathes at a discount for you if
you ever need one. The Harrison lathes are good quality of course, but I
have been pleasantly surprised at the 14x40 Nardini that we have at our
office. We have a small tool room and the home office let me have a new
Nardini. I would have loved a Harrison Alpha but for what we do the $$
don't make sense. The Nardini is a better machine than I expected,
cetainly better than most although not quite as good as the Harrison
manuals. I just can't get used to the "lay-out" of the thing. Small
complaint I guess.
Anyway thanks for bouncing this around with me. I've got a couple more
ideas as a result.
The feedrate is determind by pressure, same as flat
That sends up a red flag. I'm sure you've not talking about lots of
pressure, but I can't help but wonder if it's more than the matrix of the
wheel can withstand (as you alluded, maybe a bad wheel).. How are the CBN
wheels bonded? I've never seen one.
I'd still be inclined to go with a 38A Norton wheel, but that may not work
with that type of machinery. It sure would be my choice for the level of
technology with which I'm familiar. Years ago I made three sets of
supercharger planetary components for Auburn and Cord, each of which used
the same blower guts, but a different housing (Auburn is straight 8, Cord
V8). I recall it ground beautifully, as do the vast majority of hard
To be perfectly honest, you're way out of my league. While we worked to
close tolerances, I don't recall anything tighter than .0001" (for
flatness). We used to grind a thin cover for the bearing housings of the
guidance system of the missile. They were made of A286 stainless, drilled
and countersunk, then hardened. We'd rough them on a surface grinder, but
to get them flat, we'd spin them in a fixture that, for all practical
purposes, emulated the machine you are using, but very crude and old. It
was an old B&S grinder with drip oilers, if that helps put things in
perspective. We relieved the side of the wheel, using only the outside
1/4" or so, then, with the work head and grinding heat perfectly parallel,
we'd float the part in a pocket fixture, driven by a loose fitting pin in
one of the holes. We could get the parts flat (less than .0001") by
grinding both sides. It's an excellent way to grind, because there's
nothing distorting the part, which was actually held in place by the wheel.
While it may not be what you're hoping for, it might be a good idea to hit
them on a surface grinder to make them uniform in size before using the
double disk machine. Sort of a rough grinding operation, leaving only
enough stock to bring them flat and to size. Dunno. Might be too much
Then I'd suggest to them that they change the matrix in their wheels!
Sounds like the old "pass the buck" thing to me. These wheels must be
fairly expensive, or they'd pony up another, at least to see if it, too,
chipped. I have a philosophy when it comes to such matters. If, in the
end, I find they're the problem, I make certain to use their competitors
for all future transactions.
Do you have any idea when the wheel fractures? It might not be when the
piece finally dislodges. I'm still wrestling with the pressure.
So the gage part acts like a stop, but electronically? You'll have to
forgive me, for any machine I've ever run, everything was determined by my
hand. I've never used so much as a DRO in my entire machining career.
Everything was done the hard way.
I like the idea that the part never leaves the wheel. It is likely where
the chipping occurs. By its nature, there's not really a need for the parts
to go outside the wheel boundary, so it might be an excellent thing to
pursue, especially if coolant can lift a part occasionally.
Curious. How much of the wheel face is being used? Is it like a Blanchard?
Thanks for your candor on the manual machines. I'll certainly keep your
kind offer in mind.
Not sure I did you any good, but I'd sure enjoy hearing what you conclude.
Why don't you keep us all posted? These are the kinds of problems that are
interesting, and often very difficult to solve. In the end, you'll wonder
how you missed it, assuming it's not a bad wheel, anyway.
I use a half face resperator with 3M cartridges and prefilters.
I also use this setup when cleaning up model engine castings with
a grinding stones or cratex (using a foredom).
The rockwell toolmakers grinder is not capable of flood coolent but
does have a mist coolent system. Sound like I sould install an exhaust
fan and try it out.
"Harold and Susan Vordos" wrote in
They are resin bonded. The one we are using is designed for low pressure
double disk grinding. The main advantage is you can go a long time
between dressing. So they pay for themselves in high volume applications.
We were able to go about an hour between dressing with the CBN. We were
doing 140 pcs. per load. The cycle time starts out under 30 seconds. In
an hours time the cycle time gets to be about a minute and we dress the
wheel which takes about five minutes. If the blanks were off of a cnc
rather than a multi spindle we could do better. The amount being ground
was excessive IMO.
I've got a bit of an update. A customer has comitted to the machine
today, and for his application I'll be using conventional wheels and
double disk grinding. Funny you should mention turbos...
A-286. Ick. Ive cut a bit of that. I worked on an interesting turnkey
once on a Y-axis sub spindle twin turret lathe. We rough and finish
milled then rough and finish ground an A-286 part. We were running parts
and I heard something heavy fall into the chip pan. I looked into the
machine and couldn't see anything wrong so I went around and looked in
the chip pan. There was half of a 3/4" end mill in there. Never stopped
the machine and didn't damage the part. The end mill just got dull and
broke. Tough stuff that A-286.
I'm amazed at how guys like you made parts back in the day. If you were
to watch us grind on this machine you wouldn't believe how easy it is.
That is once you get the fixture right, get a good wheel, and dial in the
program. On manual grinders you pretty much have to be on your game all
day. Now-a-days all the development is up front (my favorite part) then
once it's dialed in and proven capable there is not a lot of skill
required in running the machine.
Yeah it would be too many ops. That customer is surface grinding now.
Productivity and quality is not nearly as good as what we are doing.
They were passing the buck without a doubt. That's what prompted me to
ask you about double disk grinding. I was hoping to find another supplier
and maybe get some advice. Anyway I don't have a lot of faith in these
guys anymore. I'm going to give Norton a call for this new deal.
I'm telling you that you would find no joy in operating this machine.
Setting it up and wringing it out is where the fun is. The way the gage
works is that you set it with gage blocks and zero it out. During the
cycle the gage advances and measures the part or dummy part in process.
As the machine gets close to size it reduces the pressure (and therefor
the feed rate), then when the gage measures zero it pulls the wheel up,
Thats our current theory. If that customer wants to carry on, I'm
thinking we either change the wheel or use a fixture where the part stays
under the wheel and we use a dummy for gaging.
Sort of except the wheel is not segmented. The whole face of the wheel is
That's why talking things out is always good. It gets your mind out of
the rut it's in and suddenly you see the problem. Unfortunately I wasn't
involved in this project until recently, so I'm trying to get up to speed
fast. I'll let you know how the new deal goes, and if we ever get back to
the other jobs I'll let you know what we figure out.