Why I was looking for info about strain gauges and the like

I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well,
it doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at
some torque number but I don't know how the clutch material behaves
over time, if the pressure will change significantly over time for the
same torque. The tapping head will easily handle #2 to 1/4 inch taps
in soft and hard materials. I think it will even handle taps as small
as #0. Experiments show that I am able to use it to tap #0-80 holes in
steel without breaking taps but I am not yet convinced that a machine
tool will be able to. When I'm tapping there is some feel whereas a
CNC machine will only feed the tap at the programmed speed. I was
hoping that some sort of strain gauge applied to some sort of rotating
shaft subjected to some amount of torque might be pretty
straightforward and cheap. It is not. But thanks for the replies
anyway, I really appreciate it.
Cheers,
Eric

 

[Maynard A. Philbrook Jr.]

In article <rofk8atsb3f875hhhjauu8179qnhbjfnqf@4ax.com>,
etpm@whidbey.com says...

I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well,
it doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at
some torque number but I don't know how the clutch material behaves
over time, if the pressure will change significantly over time for the
same torque. The tapping head will easily handle #2 to 1/4 inch taps
in soft and hard materials. I think it will even handle taps as small
as #0. Experiments show that I am able to use it to tap #0-80 holes in
steel without breaking taps but I am not yet convinced that a machine
tool will be able to. When I'm tapping there is some feel whereas a
CNC machine will only feed the tap at the programmed speed. I was
hoping that some sort of strain gauge applied to some sort of rotating
shaft subjected to some amount of torque might be pretty
straightforward and cheap. It is not. But thanks for the replies
anyway, I really appreciate it.
Cheers,
Eric

I see you are dealing with small taps..

If you drive the head/holder directly via a Stepper motor system, you
can regulate maximum current to the stepper. Measure the time it takes
to complete each step.

A longer than normal time to complete a step will indicate that the
tap is getting a little used up or needs lube.

Depending on the size of the tap you're driving, you can set the
maximum motor stepping current and also need to monitor the time
it takes for the step to complete before doing another step.

Not sure how a stepper motor would work attached to existing equipment?
it may do just fine. The concept would be the same.

You could also use a servo motor but that starts to get a little
expensive.

Jamie

 

[Tim Wescott]

On Thu, 11 Dec 2014 17:05:51 -0800, etpm wrote:

I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well, it
doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at some
torque number but I don't know how the clutch material behaves over
time, if the pressure will change significantly over time for the same
torque. The tapping head will easily handle #2 to 1/4 inch taps in soft
and hard materials. I think it will even handle taps as small as #0.
Experiments show that I am able to use it to tap #0-80 holes in steel
without breaking taps but I am not yet convinced that a machine tool
will be able to. When I'm tapping there is some feel whereas a CNC
machine will only feed the tap at the programmed speed. I was hoping
that some sort of strain gauge applied to some sort of rotating shaft
subjected to some amount of torque might be pretty straightforward and
cheap. It is not. But thanks for the replies anyway, I really appreciate
it.

Hey Eric:

That changes things. I think everyone was assuming the same thing that I
was -- that you were working on a one-off, and thus that a bunch of
development work would have been downright stupid.

But hey, you're not! I think that wireless is probably still too spendy,
unless you're going to make 10,000 tapping heads. However, I think you
could make a battery-powered unit that goes on the tapping head & uses a
"naked" strain gauge, that holds the maximum torque exerted so the
operator can look at it after the fact. And I think it could all be done
for a BOM cost that's reasonable.

Now, whether we could coach you through designing all of this -- I
dunno. Several of us do consulting and could do the work, but we'd want
to get paid a fair wage for our labor, which may amount to more than you
can afford -- but again, I dunno what your budget is.

Personally, and in spite of the fact that I'd love to talk you into
having me do the design for you, I think that in the larger business
sense you may be better off to concentrate on getting your tapping head
working and on the market in the minimum amount of time. Save the strain
gauge for a year or two down the road, when you have revenue coming in
and you want to have something to differentiate you from your
competitors, and something to tempt your existing installed base to
upgrade.

--
www.wescottdesign.com

 

[Jasen Betts]

On 2014-12-12, Tim Wescott <tim@seemywebsite.com> wrote:

On Thu, 11 Dec 2014 17:05:51 -0800, etpm wrote:


Hey Eric:

That changes things. I think everyone was assuming the same thing that I
was -- that you were working on a one-off, and thus that a bunch of
development work would have been downright stupid.

But hey, you're not! I think that wireless is probably still too spendy,
unless you're going to make 10,000 tapping heads. However, I think you
could make a battery-powered unit that goes on the tapping head & uses a
"naked" strain gauge, that holds the maximum torque exerted so the
operator can look at it after the fact. And I think it could all be done
for a BOM cost that's reasonable.

If you're putting a battery in it wireless is cheap, probably cheaper
than a display. eg: bluetooth. IRDA could also work.

--
umop apisdn

 

Tim Wescott <seemywebsite@myfooter.really> wrote:

Assuming that the battery life issue could be dealt with, if he's
going to make 100,000 units, or even 10,000, then some custom
Bluetooth implementation is probably best.

You will have a rotating shaft passing through the middle of the
gadget... stick some neodymium magnets on the shaft and put some
(stationary) coils next to them. You have to deal with the RPM not
being much and that it can go either direction... maybe you use the
coils to charge a supercap or NiMH AA or something. Also you have to
keep coolant/oil and metal chips out of the (hopefully small) air gap
between the magnets and the coils. For tapping, say, 0.25" or M6 or
bigger, the generator probably won't mess with your torque readings, but
for smaller holes it might.

If a Bluetooth phone earpieces work OK in a machine shop, then this
probably isn't a big deal, but: some shops might be a little noisy at
RF. Mostly arc welding, but perhaps also noise from the control
computers, stepper motors, etc.

At a previous job, I used a Haas mill. It had encoders on all the axes,
but if you wanted to get really precise, it had a probe from (I think)
Renishaw to help. Half of it went in the chuck where the mill bit would
normally go - that half had a battery, a bunch of IR LEDs in a circle,
and a spring-loaded plunger sticking out of the end. The other half was
a box with an IR reciever that bolted to the (fixed) splash shield on
the mill and had a wire to the mill's computer. If I remember right,
you basically told the mill computer to listen for the probe, and then
used the handwheels to move the mill head until the plunger juuust
contacted the surface. That would send a signal via IR, and the mill
computer would accept that as the new zero / starting point.

I don't know what the battery life for that probe was like, but driving
all those IR LEDs has to take a little juice. It wasn't a commercial
production shop, so we didn't use the probe that often. I get the
impression that a probe like that would get used less often than a
tapping head, though.

Matt Roberds

 

[Mark Zenier]

In article <rofk8atsb3f875hhhjauu8179qnhbjfnqf@4ax.com>,
<etpm@whidbey.com> wrote:

I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well,
it doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at
some torque number but I don't know how the clutch material behaves
over time, if the pressure will change significantly over time for the
same torque. The tapping head will easily handle #2 to 1/4 inch taps
in soft and hard materials. I think it will even handle taps as small
as #0. Experiments show that I am able to use it to tap #0-80 holes in
steel without breaking taps but I am not yet convinced that a machine
tool will be able to. When I'm tapping there is some feel whereas a
CNC machine will only feed the tap at the programmed speed. I was
hoping that some sort of strain gauge applied to some sort of rotating
shaft subjected to some amount of torque might be pretty
straightforward and cheap. It is not. But thanks for the replies
anyway, I really appreciate it.

What about a shaft encoder at both ends of a torsion shaft?
The torque showing up as the difference in angular position.

And it would have the advantage that the electronics could remain
stationary.

Mark Zenier mzenier@eskimo.com
Googleproofaddress(account:mzenier provider:eskimo domain:com)

 

[George Herold]

On Saturday, December 13, 2014 12:27:50 PM UTC-5, Mark Zenier wrote:

In article <rofk8atsb3f875hhhjauu8179qnhbjfnqf@4ax.com>,
etpm@whidbey.com> wrote:
I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well,
it doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at
some torque number but I don't know how the clutch material behaves
over time, if the pressure will change significantly over time for the
same torque. The tapping head will easily handle #2 to 1/4 inch taps
in soft and hard materials. I think it will even handle taps as small
as #0. Experiments show that I am able to use it to tap #0-80 holes in
steel without breaking taps but I am not yet convinced that a machine
tool will be able to. When I'm tapping there is some feel whereas a
CNC machine will only feed the tap at the programmed speed. I was
hoping that some sort of strain gauge applied to some sort of rotating
shaft subjected to some amount of torque might be pretty
straightforward and cheap. It is not. But thanks for the replies
anyway, I really appreciate it.

What about a shaft encoder at both ends of a torsion shaft?
The torque showing up as the difference in angular position.

And it would have the advantage that the electronics could remain
stationary.

That's what I was thinking, it could just be two magnets, or optical.

George H.

Mark Zenier mzenier@eskimo.com
Googleproofaddress(account:mzenier provider:eskimo domain:com)

 

[George Herold]

On Saturday, December 13, 2014 12:52:14 PM UTC-5, et...@whidbey.com wrote:

On Fri, 12 Dec 2014 21:35:27 GMT, mzenier@eskimo.com (Mark Zenier)
wrote:

In article <rofk8atsb3f875hhhjauu8179qnhbjfnqf@4ax.com>,
etpm@whidbey.com> wrote:
I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well,
it doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at
some torque number but I don't know how the clutch material behaves
over time, if the pressure will change significantly over time for the
same torque. The tapping head will easily handle #2 to 1/4 inch taps
in soft and hard materials. I think it will even handle taps as small
as #0. Experiments show that I am able to use it to tap #0-80 holes in
steel without breaking taps but I am not yet convinced that a machine
tool will be able to. When I'm tapping there is some feel whereas a
CNC machine will only feed the tap at the programmed speed. I was
hoping that some sort of strain gauge applied to some sort of rotating
shaft subjected to some amount of torque might be pretty
straightforward and cheap. It is not. But thanks for the replies
anyway, I really appreciate it.

What about a shaft encoder at both ends of a torsion shaft?
The torque showing up as the difference in angular position.

And it would have the advantage that the electronics could remain
stationary.

Mark Zenier mzenier@eskimo.com
Googleproofaddress(account:mzenier provider:eskimo domain:com)
I have decided to use a battery for at least the alarm part because no
power would be drawn until the head detects a fault. I thought about a
generator but it seems to me that adds too much complexity, at least
in the quantities I might be able to sell. I don't think encoders
would work because I don't think the shaft would twist enough when
tapping very small holes. But this is just seat of the pants figuring.
I could make a setup on the lathe to test your idea. I have encoders
with 2500 lines which means that read in quadrature there would be
10,000 pulses. And maybe the shaft would twist up 1/10,000 of a
revolution when tapping 2-56 holes.
Thanks,
Eric

Hi Eric, Marks idea includes a torque shaft..which I assume is a section that's designed to twist more with the torque.. (Do they make such things?)

George H.

 

On Thu, 11 Dec 2014 23:36:58 -0600, Tim Wescott <tim@seemywebsite.com>
wrote:

On Thu, 11 Dec 2014 17:05:51 -0800, etpm wrote:

I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well, it
doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at some
torque number but I don't know how the clutch material behaves over
time, if the pressure will change significantly over time for the same
torque. The tapping head will easily handle #2 to 1/4 inch taps in soft
and hard materials. I think it will even handle taps as small as #0.
Experiments show that I am able to use it to tap #0-80 holes in steel
without breaking taps but I am not yet convinced that a machine tool
will be able to. When I'm tapping there is some feel whereas a CNC
machine will only feed the tap at the programmed speed. I was hoping
that some sort of strain gauge applied to some sort of rotating shaft
subjected to some amount of torque might be pretty straightforward and
cheap. It is not. But thanks for the replies anyway, I really appreciate
it.

Hey Eric:

That changes things. I think everyone was assuming the same thing that I
was -- that you were working on a one-off, and thus that a bunch of
development work would have been downright stupid.

But hey, you're not! I think that wireless is probably still too spendy,
unless you're going to make 10,000 tapping heads. However, I think you
could make a battery-powered unit that goes on the tapping head & uses a
"naked" strain gauge, that holds the maximum torque exerted so the
operator can look at it after the fact. And I think it could all be done
for a BOM cost that's reasonable.

Now, whether we could coach you through designing all of this -- I
dunno. Several of us do consulting and could do the work, but we'd want
to get paid a fair wage for our labor, which may amount to more than you
can afford -- but again, I dunno what your budget is.

Personally, and in spite of the fact that I'd love to talk you into
having me do the design for you, I think that in the larger business
sense you may be better off to concentrate on getting your tapping head
working and on the market in the minimum amount of time. Save the strain
gauge for a year or two down the road, when you have revenue coming in
and you want to have something to differentiate you from your
competitors, and something to tempt your existing installed base to
upgrade.

Greetings Tim,
Yeah, mission creep can be a bad thing. I do already have a few tiny
transmitters that can be had pretty cheap. If I only want to send a
string of pulses they would work. I already am planning on using a
transmitter to send a signal to an alarm box outside of the machine to
let the operator know when a hole doesn't get tapped. I have
breadboarded the electronics and they work fine on the bench. If I do
add some sort of torque sensor it will need to be after I have
produced the first model. So I can afford to pay someone to design the
electronics if need be.
Thanks,
Eric

 

On Fri, 12 Dec 2014 21:35:27 GMT, mzenier@eskimo.com (Mark Zenier)
wrote:

In article <rofk8atsb3f875hhhjauu8179qnhbjfnqf@4ax.com>,
etpm@whidbey.com> wrote:
I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well,
it doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at
some torque number but I don't know how the clutch material behaves
over time, if the pressure will change significantly over time for the
same torque. The tapping head will easily handle #2 to 1/4 inch taps
in soft and hard materials. I think it will even handle taps as small
as #0. Experiments show that I am able to use it to tap #0-80 holes in
steel without breaking taps but I am not yet convinced that a machine
tool will be able to. When I'm tapping there is some feel whereas a
CNC machine will only feed the tap at the programmed speed. I was
hoping that some sort of strain gauge applied to some sort of rotating
shaft subjected to some amount of torque might be pretty
straightforward and cheap. It is not. But thanks for the replies
anyway, I really appreciate it.

What about a shaft encoder at both ends of a torsion shaft?
The torque showing up as the difference in angular position.

And it would have the advantage that the electronics could remain
stationary.

Mark Zenier mzenier@eskimo.com
Googleproofaddress(account:mzenier provider:eskimo domain:com)

I have decided to use a battery for at least the alarm part because no
power would be drawn until the head detects a fault. I thought about a
generator but it seems to me that adds too much complexity, at least
in the quantities I might be able to sell. I don't think encoders
would work because I don't think the shaft would twist enough when
tapping very small holes. But this is just seat of the pants figuring.
I could make a setup on the lathe to test your idea. I have encoders
with 2500 lines which means that read in quadrature there would be
10,000 pulses. And maybe the shaft would twist up 1/10,000 of a
revolution when tapping 2-56 holes.
Thanks,
Eric

 

On Sat, 13 Dec 2014 10:37:36 -0800 (PST), George Herold
<gherold@teachspin.com> wrote:

On Saturday, December 13, 2014 12:52:14 PM UTC-5, et...@whidbey.com wrote:
On Fri, 12 Dec 2014 21:35:27 GMT, mzenier@eskimo.com (Mark Zenier)
wrote:

In article <rofk8atsb3f875hhhjauu8179qnhbjfnqf@4ax.com>,
etpm@whidbey.com> wrote:
I have been designing a new type of tapping head aimed at small shops
with older equipment. It is designed to keep taps from breaking, it
detects missing, short or undersize holes, and it does windows. Well,
it doesn't really do windows. Anyway, I thought that as long as I was
making something that did all that it would be neat if it could report
on dull taps too. I may be able to use clutch pressure to arrive at
some torque number but I don't know how the clutch material behaves
over time, if the pressure will change significantly over time for the
same torque. The tapping head will easily handle #2 to 1/4 inch taps
in soft and hard materials. I think it will even handle taps as small
as #0. Experiments show that I am able to use it to tap #0-80 holes in
steel without breaking taps but I am not yet convinced that a machine
tool will be able to. When I'm tapping there is some feel whereas a
CNC machine will only feed the tap at the programmed speed. I was
hoping that some sort of strain gauge applied to some sort of rotating
shaft subjected to some amount of torque might be pretty
straightforward and cheap. It is not. But thanks for the replies
anyway, I really appreciate it.

What about a shaft encoder at both ends of a torsion shaft?
The torque showing up as the difference in angular position.

And it would have the advantage that the electronics could remain
stationary.

Mark Zenier mzenier@eskimo.com
Googleproofaddress(account:mzenier provider:eskimo domain:com)
I have decided to use a battery for at least the alarm part because no
power would be drawn until the head detects a fault. I thought about a
generator but it seems to me that adds too much complexity, at least
in the quantities I might be able to sell. I don't think encoders
would work because I don't think the shaft would twist enough when
tapping very small holes. But this is just seat of the pants figuring.
I could make a setup on the lathe to test your idea. I have encoders
with 2500 lines which means that read in quadrature there would be
10,000 pulses. And maybe the shaft would twist up 1/10,000 of a
revolution when tapping 2-56 holes.
Thanks,
Eric

Hi Eric, Marks idea includes a torque shaft..which I assume is a section that's designed to twist more with the torque.. (Do they make such things?)

George H.

If I was designing a device for tapping only a small range of sizes
then a "torque shaft" could be used. But the head must tap between #2
and 1/4. So a shaft that would have much deflection with a #2 tap
would probably just twist off with a 1/4 tap.
Eric