I bought a VERY old heat treat oven quite a long while back...
I'd like to check that it is reasonably calibrated before heat
treating this part. Is there some sort of thermocouple and high temp
wire I can buy and take a voltage reading? I only know of this
concept, not any specifics.
I made a custom bolt out of 4340 for my 30 ton shop press. Looks like
I should heat to 1525 and quench in oil. Then temper to 450. correct?
I'll use SHMBO oven for tempering (when she goes shopping)
Karl
The accurate, inexpensive way is to buy a second-hand temperature
controller and some high-temp Type K thermocouple wire.
formatting link
Be sure you can find the controller's operating manual on line. They
can be very tricky to set.
You could put a relay in a box with the controller and plug the oven
into it. The controller's PID function reduces power as the
temperature approaches the set point, so the oven doesn't overheat.
I have similar thermocouples from a different local supplier. Don't
try to cheat on the temperature rating. Cheaper 1200F glass braid
rapidly turns to dust in a candle flame at 1500F.
There are also hand-held battery powered thermocouple thermometers
like my Fluke 52 but they don't usually go cheap.
jsw
I expect he would be far better off getting a metal sheathed
thermocouple with the wire connections well away from the hot end. They
can be had in various alloys and good to 1100C or more depending on
alloy and the hot end can be poked in while the wire connections can be
at ambient so don't need to be anything special. They are available in
many lengths and diameter in that configuration. I had a quick look at
Omega but they had so many options I though it was pointless to point
out one when the OP can chose one to suit his needs. They are not
expensive although I thought a few I looked at on the Omega site were
and they were in the US, I'm in the UK and thought the prices expensive.
I have various CAL temperature controllers and use them with K and N
type thermocouples, I also have a DVM with K type input and they all
tally very well with one another.
Well ... Omega makes a series of thermocouples designed to
handle a range of temperatures, and they make both readout boxes for
them or very nice controllers which will work with the thermocouple to
produce the desired temperature.
I picked up a heat treat furnace with no controller at a tool
flea market some years ago. I already had a nice Omega controller from
an overstock for a project that another poster on this newsgroup had
left over. He was asking $100.00 for unused ones which sold new for
$200.00, so I got one. I already had some Type-K thermocouples in
stainless steel housings which I could use there.
The controller can be told to work in either degrees C or
Degrees F -- whichever works for your project.
It will pause at about the 2/3 or 3/4 point on the way up to
measure how much overshoot there is and compensate for it. I set 1800F
as my target, and I got a total overshoot of 1 degree F when it reached
the set point.
Note that the output voltage of a thermocouple is non-linear,
and a *very* low voltage compared to most multimeters. I think that
somewhere in the range of 35 mV is a reasonable value. And to get
absolute readings you need to compare the output of the thermocouple
with a standard (in the old days, a second thermocouple of the same type
dipped in a container of water and ice for a 0 C (32 F) reference.
There are now electronic reference junctions built into such
controllers.
The controller I got had the ability to directly control fairly
low current, or to put out a 5VDC signal. I used that 5V signal to
control a solid state relay (Crydom 24025 IIRC, capable of switching 240
VAC at up to 25 Amps), and let *that* control the heating element in the
oven.
The controller is a proportional controller -- it sits with the
heating element on until it gets within a certain range (determined by
the test on the way up), and then starts turning the element on for
perhaps 90% of a 10 second cycle, then 80, then 70, on down until it
stabilizes at some percentage on time which maintains the target
temperature.
The same oven and controller can be used for the lower
temperature tempering if you first let it cool off enough. I've done it
that way with some D2 air hardening steel.
I won't comment on the tempering because I don't know how hard
you need it to be.
The first site I checked ======================================================================
Heat treatment for strengthening is done at 1525 F followed by
an oil quench. For high strength (over 200 ksi) the alloy should
first be normalized at 1650 F prior to heat treatment. See
"Tempering" for strength levels.
======================================================================
So the first question is whether you need over 200 ksi strength? I
don't know what a 30 ton shop press needs, and I suspect that it varies
a lot with design.
The same site says (in the "Tempering" section):
======================================================================
The temperature for tempering depends upon the strength level
desired. Before tempering the alloy should be in the heat
treated or normalized & heat treated condition - see "Heat
Treatment". For strength levels in the 260 - 280 ksi range
temper at 450 F. For strength in the 125 - 200 ksi range temper
at 950 F. Do NOT temper the alloy if it is in the 220 - 260 ksi
strength range as tempering can result in degradation of impact
resistance for this level of strength.
======================================================================
So your 450 F tempering suggests that you want to be in the 260-280 ksi
range, which also suggests that you need to normalize it first.
And how long you hold it at temperature is a function of the
thickest part -- with (IIRC) about 60 minutes per inch of thickness.
The site which I've quoted above is:
Why? Turn off the oven while you quench, and wait until it
cools off enough to set to the lower temperature. Bound to be more
precise than the typical home oven.
FWIW -- when I built the controller housing, I included a small
switch to open the wire from the controller to the SSR (Solid State
Relay), so I could be sure that it was not heating, but could use the
thermocouple to monitor the temperature in the oven. Note that if you
leave the door open while quite hot, the corrosion of the heating
elements will increase, and when you think that it is cool enough and
close the door, the temperature will shoot back up quite a bit. Better
to just be patient. It takes my oven (quite small, and 120 VAC
operation) about an hour to make it up to 1800 F, and longer to make it
back down to reasonable temperatures.
Also -- note that the dimensions of hardened steel are larger
than those of the annealed steel, so make sure that your thread and
shank diameters are made with this under consideration. You certainly
won't be able to use a die to fix the size after hardening. :-)
If this is an ordinary bolt in a reasonable size, what is wrong
with buying one of the proper size and strength off the shelf, instead
of going to all this work? Or do you want to get some experience at the
hardening process?
Enjoy,
Don.
"David Billington" wrote
I thought about suggesting a probe too. The wire could be snaked in
around the insulation while a probe would need a hole or thermowell,
drilled from outside and avoiding any heating elements. The
thermocouples in my woodstove are wires and sometimes I move them
around to see if another placement gives a better reading. If the tool
being heat-treated is in a closed box with charcoal only flexible t/c
wires work well.
These panel connectors fit into modular audio/video wall plates with
only a little whittling:
formatting link
have wood stove and cooking pot thermocouples, phone, TV antenna,
and two speaker (solar panel) banana jacks in the one behind the
computer.
jsw
Not sure it makes any difference as I have used probes that were 1mm
diameter and could be bent to fit but the same applies to bare thermo
couple wire and junction . I think the off the shelf probes are more
suited to this application of poking in through insulation. I assume we
are talking about the same sort of beast , a metal sheathed probe
covering a thermocouple junction.
Shoot, I didn't allow for it growing. The only way I had to test the
1.5 by 16TPI threads was to put the whole 30 ton cylinder in the lathe
and try to thread it on. I ended up a bit tight. Took fiddling with
the old triangle file but finnally got it to spin all the way in real
nice.
This bolt holds the block on the end of the cylinder to bolt press
beak and other dies to. I just wanted it as good as i could get it. I
guess "as is" is good enough. I've never trusted my old heat treat
oven, guess it won't get used again this time.
Karl
At 2000F a type K is giving you about 50 millivolts and you can read
that on an ordinary DVM. But a chinese thermocouple meter from ebay is
cheap and easy and should be accurate enough for your needs.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.