Sciencemadness Discussion Board

design choice for electrical resistance chemistry lab furnace

coppercone - 5-5-2018 at 14:38

I want to make a lab furnace. It will use Kanthal A1 wire.

I have a precision 6kW DC supply (yes its over the top) to use as a power source, and a PID controller, the PID gives a 4-20mA or 0-10V signal to control the DC supply (i think it has PWM too).

I want to make it with a upward swinging door (so you don't have a wall of hot shit ready to burn you swinging on some hinges), a interlock to disable power if opened, possibly a circuit to measure coil impedance integrity periodically

I am a EE and I don't like triac control schemes etc. Prefer my nice supply.

Major questions:

1) How much of a contamination concern is bare wire in the furnace?

I know there is the option of putting coils of wire into ceramic tubes. Does this significantly stop off gassing? Is off-gassing of the heating elements a concern to high temperature chemistry? I assume I would put a sealed bomb flask made of quartz or a crucible with a lid on it in the furnace.

2) Do any special precautions need to be taken for inert gas introduction? I was just going to drill a hole in the top middle of the furnace and allow for gas entry (argon). Does it need some kind of diffuser and do they build furnaces that have inert gas inlets with some kind of special interior airflow considerations?

3) I see that most lab furnaces (in relation to question 1) just have the bear coils exposed, but in grooves. However, I saw one more advanced design that has the coil behind a kind of ledge made of the ceramic liner, much like 'mood lighting' that you would hide behind home interior decorative molding. Is there a significant need to do this to prevent the formation of thermal hot spots due to direct radiation emitted by the heating coils (they will be much brighter then the brick and more radiant). https://www.youtube.com/watch?v=1zYmjtr0akk this video shows this type of furnace (1 minute 30 seconds in). ??

4)Are there any other special considerations for research-chemistry laboratory furnaces that are not covered by guidelines for design of so called laboratory furnaces (which seem to be more used for things like heat treatment, materials testing ,etc). I mean physical design specifications. I am not that interested in software specific considerations like ramp times, cycling times, etc... these can be programmed in at a later date (i can do things like custom PID control on microcontroller if necessary to do something that an off the self PID control module does not have)

I am also aware that fluid motion inside of the furnace will strip the elements so they should be run at a lower power when shielding gas is used.

I thought about adding things like ports for the introduction of multiple gasses (in addition to the inert gas), but I think a tube furnace is a better choice for 'continuous process' type reactions.

[Edited on 5-5-2018 by coppercone]

[Edited on 5-5-2018 by coppercone]

zed - 8-5-2018 at 14:46

https://www.youtube.com/watch?v=M5oKvHktmNU&t=1s

--------------------------------------------------------------------


“Do you wish a ticket to go away and come back or do you wish a ticket to go away and never come back?” the ticket agent asked wiping sleep out of his eyes.

“We wish a ticket to ride where the railroad tracks run off into the sky and never come back—send us far as the railroad rails go and then forty ways farther yet,” was the reply of Gimme the Ax.

“So far? So early? So soon?” asked the ticket agent wiping more sleep out his eyes. “Then I will give you a new ticket. It blew in. It is a long slick yellow leather slab ticket with a blue spanch across it.”

----------------------------------------------------------------------

Lots of furnace designs out there. Youtube is hosting quite a few.

Don't know what you are actually planning. How big? How hot? Will a Jewelers Burnout Furnace suffice?

Vogel has a design for a simple tube-furnace.

And then, there is this guy. https://www.youtube.com/watch?v=AyJAwBQ2Iuo

Now, I can do everything he does. But, it would take me ages, and the product would be crude.



[Edited on 8-5-2018 by zed]

[Edited on 8-5-2018 by zed]

coppercone - 8-5-2018 at 18:01

Mine will have interior dimensions of approximatly 7.5 x 7.5 inches.

It will go to the limit of the kanthal A1 wire, which is 1300C in this application.

It will have a maximum wall load of 4.48kW

The maximum element surface load per square centimeter is 2488 cm^2/watt

The inner surface will use the dense firebrick (not the super foamy stuff) that will be more resistant to cracking, at the expense of thermal escape, I would rather insulate the outside better then risk having to do a furnace repair, I read the soft stuff is much less durable.

I don't know how im gonna insulate it exactly, I have plenty of rockwool but I may add a secondary layer of foamy firebrick (insulative) type over the hard interior.

I will use a wire between 2 to 3 mm inserted into grooves in coils. I determined that using 4 channels going across the furnace with kanthal A1 wire will allow for operation. I found that the surface area of the wound coils will exceed the minimum limit required for 1300C operation by a significant margin.

I have a water cutting saw and welding equipment, electrical test equipment, electronics test equipment, etc.

I watched that video. It is for making knives. I am somewhat interested in heat treating parts but again I would like to approach this from a chemistry lab standpoint, hence my questions.

elementcollector1 - 8-5-2018 at 18:22

0) What's the current going through the wires? Voltage? What's the gauge of the kanthal wires (and, asking for a friend, where'd you get them from? ;) )

1) Not much, unless you've got stuff touching the wires (diffusion might technically occur at these higher temperatures.

2) As long as you've got the takeoff on the opposite side of the lead-in (with as much space between them as possible), and there's enough mass flow for the air to be pushed out, you should be fine.

3) I'd imagine this is for slow, even heating. Thermodynamically, it's not good for your furnace, as it introduces a 'barrier' of thermal mass that the coils essentially have to get past, compared to the standard design which allows radiant heat flux from the coils in the walls to directly interact with the sample in the furnace. They're typically placed in grooves so that the radiant heat in all the other directions is 'stored' by the ceramic walls, allowing the furnace to retain heat for longer when, say, the door's opened.

4) Not that I can immediately think of. Your best bet is to pay attention to the purity and quality of the materials you're using, as with any research application.


Quote:
I found that the surface area of the wound coils will exceed the minimum limit required for 1300C operation by a significant margin.


What's your metric/equation for relating surface area to operating temperature? I'm curious, as I've never heard of this relationship before and want to build a furnace of my own.

coppercone - 9-5-2018 at 06:01

To do the calculation you use the kanthal furnace book.

Wall load is determined by your furnace surface area. There is a lookup table you use to see the theoretical wattage.

You can misuse this calculator as it tells you the spacing and total wire lenght.
http://hamwaves.com/antennas/inductance.html

Basically you put in the diameter of the coil and wire, put some kind of arbitrary lenght (i.e. 100mm) and vary the number of turns till you get a correct spacing. Then you have a base unit of coil wire lenght.

Then you take your furnace and imagine the groove pattern that you cut into it and estimate the total lenght of trench you have.

Then you use the base unit generated by thr calculator to determine how much the total wire lenght is of your coil sitting in the trench.

Then you take the surface area of the theoretical wire cylinder and compare it to your loading number gathered from the lookup table.

Using thecalculator you can run the numbers for various wire gauges.

The hand book is here
https://www.google.com/url?sa=t&source=web&rct=j&...

Pages 6 and 7 will clue you in. Your coil is between 5 to 8 times the diameter and the spacing is between 2 to 3 times the diameter. Use this with the ham calculator to estimate wire lenght.

The cheapest wire i found is


https://www.euclids.com/index.php?item_id=WK12X

Not ideal you probobly want 3mm like the hand book but that stuff is cheap.

You can estimate the dimensions with density and diameter. According to my calculations 1 lb is adequate

coppercone - 9-5-2018 at 10:54

To do the resistance calculations you use the resistance formula,

so you get the area of your wire (table for AWG is nice), the resistance of the element (1.45+fudge factor if you want in the handbook page 26 or something) and the length.

I basically determined what is possible with the geometry I had in mind, then rearranged the resistance equation to solve for resistance (i found it in the form to find length) with the length of coils that I found reasonable for my furnace geometry.

I got something like 10 ohms for my furnace design with 20 meters of 12awg wire that would need 203V to get the 4600 watts from it.
A bit dangerous because a short circuit of the supply will give 230V, so it can damage the furnace.

The saftey interlocks on these are kind of complicated because you can't really put some kind of low tech reliable thing like a thermoswtich on something that goes to 1400C (or can you? do things like this exist), so you need to rely on a thermocouple for monitoring if you do decide to make a independent monitoring circuit that would flip a contactor off.

I would basically advise you to use two thermocouples, one connected to a dedicated circuit that is basically set to trip at a certain temperature to prevent a meltdown, and disable the power supply by opening a contactor, and another thermocouple to interface with your control system.

The controller you use (PID etc) will be much less reliable then a properly designed safety interlock circuit, assuming you are capable of building it (20$ pid controllers people use do not inspire confidence). This circuit should only have one function, to latch when a particular temperature is met for some period of time (to prevent noise) and turn it off. Should be independently powered and heavily protected.

I may design something and post a schematic.

I can't help you with triac control because I am blessed to have some beefy expensive DC supplies I have no use for. I don't like working with triacs, diacs, etc, unless its small signal applications. Unless you buy a controlled you need to put a MOV, fuse, gas discharge tube, line filter, etc.

[Edited on 9-5-2018 by coppercone]

[Edited on 9-5-2018 by coppercone]

[Edited on 9-5-2018 by coppercone]

coppercone - 9-5-2018 at 11:20

Also an option is to make some kinda thermal resistor and categorize its voltage drop, lead it outside of the furnace, and connect it to a mechanical thermal switch.

I don't know how to make this reliable or how to get reliable calculations on it.

http://www.wolfautomation.com/media/pdf/tempcontrol/selco/th...

https://www.webstaurantstore.com/hatco-02-16-131-00-thermost...

Perhaps someone that knows thermodynamics and have some practical experience with high temperature insulators and stuff can tell us about the stability of using that with some kind of thermal resistor/'voltage divider'

But you would need to establish thermal time constants and how fast the furnace will get damaged and how repeatable it will be with ware and tare.

I personally want to do PID controller + secondary thermocouple



It will probably depend on whats in the furnace too, unless its setup just right. Would love to know how easy this is.

[Edited on 9-5-2018 by coppercone]

zed - 9-5-2018 at 12:20

Knives? I'm looking at the furnace itself, not what is being done with it.

As for a door that swings upwards... To me, this seems like a dangerous idea.

Might deflect the extreme heat, from your freshly opened oven, right into your hands or face.

coppercone - 9-5-2018 at 13:57

by swinging it open you have twice the area of hot stuff facing you?

I guess I can see a problem if you swing it open like a mad man , standing directly in front of it, so the hot gasses can move towards you, but this is retard behavior

what about knives now? I am just saying that something for heat treatment might not have had as much thought put into it as a piece of lab equipment. All the people on the internet making these are basically looking to temper metal.

I just noticed that more often then not, scientific equipment has features considered standard by researchers that is not present on the low-tech industry versions of the same equipment.

[Edited on 9-5-2018 by coppercone]

coppercone - 9-5-2018 at 14:10

these are the nicest I was able to find:
https://www.youtube.com/watch?v=B-9EvUZ10_M

zed - 10-5-2018 at 14:32

OK. Seems like you may have very sophisticated goals.

Growing crystals? Fabricating prototype semiconductor devices?

Most of us, would be happy with a nice HOT functioning furnace. You seem to require more.

Now, the inert gas you want to use. And, inert to what?

Is Argon the best choice? Seems we have had discussions here, in the past, suggesting that Argon does not transmit heat well.

This of course, is an after-thought.

[Edited on 10-5-2018 by zed]

coppercone - 10-5-2018 at 15:36

I am not sure. It is just that, with these kind of home designs, given how much time you put into making it, and how much of a markup lab devices have (Sometimes 6+ times material cost), its worthwhile to add the extra features.

I have a strong electronics background.

I don't know exactly what to do with it lol. I was hoping that people would say well 'top of the line lab furnaces from some company I never heard of make this for 10,000$ and it does this this and that' and I could draw the line between build practicality, design time, resource cost and my interest in the features.

As for inert gas I have argon, helium, nitrogen in tank form, along with hydrogen. I am kind of interested in a hydrogen furnace because I want to treat mu-metal. Hydrogen should be OK but you basically need to baby it the whole time to make sure the exhaust is being burned.. though if the flow rate is low, and its outside, even if it goes out your not gonna have a catastrophe unless the tank or regulator fails.

Not even sure why you would use inert gas for chemistry materials preparation.

An example synthesis I can think of is making sodium sulfate by keeping a reactor in dull red heat for a while... but this is basic.

I don't know if I want to fabricate semiconductors (I am sure the quality you would achieve is low, fabs are expensive to run, and you can buy very beautiful transistors for cheap, but I would be interested in advanced material coatings maybe?

What is the realm of crystal growth for? Can you make useful custom laser optics or grow some kind of unique materials?

Maybe making it vacuum capable would be a good goal? I figure the outer box could be silver brazed shut (this I think will provide a easy tight 'weld'),

Sealing off the wires without doing something advanced would be difficult though... unless you can get away with using silicone, I imagine the hot furnace would have significant evaporation of the coils.

[Edited on 10-5-2018 by coppercone]

zed - 11-5-2018 at 14:43

This might be a case of overthink.

Regardless of design, there will always be problems.

Quite possibly, totally unforeseen problems.

Just gotta confront them as they come over your horizon.

The "Edison-ian" method, is to make a furnace, find it unsatisfactory.... and make a better one.

As for the cost of scientific equipment.... This stuff is wildly, unreasonably expensive.

Ten thousand dollar... "toilet seat" expensive.

Priced for unsupervised, government/academic consumption.

https://www.sigmaaldrich.com/catalog/product/aldrich/z256307...

A perfectly ordinary Zip-tie applicator, for only $340.00!

An item usually available in the 10 to 20 Dollar range.


[Edited on 11-5-2018 by zed]

[Edited on 11-5-2018 by zed]

coppercone - 11-5-2018 at 17:10

actually LOL i own one, you can buy them cheap on ebay, I got mine for like $50.

I used the cheap ones, they kinda suck in terms of ergonomics (the molded plastic ones). Mine has a nice rubber handle and the adjustment is done with a nice thumb wheel on the back, rather then a shitty little spin wheel on the side of it...

Mine is panduit brand though, that one looks kind of shitty.

If you are making cable harnesses for a living, its worth while to have (they even have full auto ones for like 8000$, that work kinda like a nail gun, or pnumatic ones for like 800 that self tighten). These things avoid carpel tunnel syndrome.

This is a nice one:
https://www.ebay.com/itm/Panduit-GTS-E-Ergonomic-Hand-Operat...

About 150$-200$ new...

I have alot of weird expensive shit though, pliers, proxxon tools, etc :P

https://www.ebay.com/p/Panduit-PTS-Pneumatic-Cable-Tie-Hand-...

You might depress the handle 4000 times doing a weeks worth of work making harnesses.

But I think they probably run like 7-8 times manufacturing cost rather then the usually accepted 4...

I had a peak at some rather obscenely priced specialty pnumatics tools, and I had a hard time believing it cost 3000$ rather then maybe 1800$.....


https://www.youtube.com/watch?v=eObc8SYGmmo

[Edited on 12-5-2018 by coppercone]

zed - 13-5-2018 at 14:19

Heh, heh.

Aldrich is primarily a Chemical supplier. So.... That cable-tie device, in the standard lab environment, would be used only a few times per Month/Year, for odd things like securing a rubber septum to a piece of glassware.

As for furnace recommendations. I'm sure someone here at ScienceMadness, has inside information on nifty furnaces, but we are currently posting out in the Boondocks, so to speak.

Technochemistry is kind of "off the beaten path".

If you try posting a query in General Chemistry, you might get more replies.

coppercone - 14-5-2018 at 08:28

Well i made some headway. With my interior dimensions

Of 7.75x7.75x7.325 and interior area of 0.224 sq meters according to the handbook my maximum wall load is

4.48kW

The bricks for the interior surface are from lowes, the hard dense durable brick. The insulation brick will be the second layer.

I choose to use 12 gauge wire, which is 2.03mm

The electrical calculations people were interested in are as follows:

Ohms mm^2/meter=1.45*fudge factor=1.508 (accounts for temp change of wire at max temp)

Cross sectional area of 12 gauge wire = 3.31

Line power = 230v

Wattage maximum at max temp = 4.48kw

P=iv

4480=230i
I=19.47a

P=i^2r
R= 11.81 ohms

Formula for wire lenght

Lenght= (resistance required*csa)/resistance area factor in begining of post

So

(11.8*3.31)/1.508=25.9 meters

So the question becomes how to fit 25.9 meters of wire in your furnace and to verify the watt loading.

I believe i iteratively came up with the passable solution you need to renember your watts of dissipation per unit lenght factor

The maximum dissipation per area is given in handbook for 3mm wire. In my case it is 1.8w/cm^2

So your bound that you should meet or exceede is

4480/1.8=2488 cm^2

So we know to get our target dissipation at 230v we need 25.9 meters of wire.

But here begin the problems and i need to work it out cuz watt loading is severely exceeded


coppercone - 14-5-2018 at 08:39

You need 39 meters of wire to satisfy the watt surface emission requirement. Is this possible? Can you cram that much fucking coiled wire in this mother fucker while maintaining design specifications?

Coppercone will soon find out

[Edited on 14-5-2018 by coppercone]

coppercone - 14-5-2018 at 09:02

So now you back solve the previous equation

39=(x*3.31)/1.508
58.812/3.31 = 17.76 ohms

So 4480=i^2*17.76

4480/17.76
Sqrt252= 15.87
P=iv

4480/15.87 =282v @ 17a yikes

Say we divide in parallel

17.76/2=8.88 ohms
2240=i28.88
I = 15.87


2240/15.87= 141 v
So 32a@141v

That can be done.

if we leave it in series you get 2950w

and a very nice figure for surface loading

Now to find the most dense geometry allowed by handbook.

The dimensions are
D=five to 8 times wire diameter
S=2 to 3 times wire diameter

So 5*2.03 =1.15cm
S= 4.06

Using the calculator i found that i can cram 908mm of wire per 100 mm of coil length.

So 10cm=90.8 cm

So i need 420cm of trenches to accommodate 39 meters of wire

Now i need to draw out my furnace geometry and determine practical bend radius for the coils and do a path length integral to determine the maximum reasonable trench area i can fit in the furnace to accommodate a coil wound to maximum specifications



[Edited on 14-5-2018 by coppercone]

coppercone - 14-5-2018 at 11:09

A preliminary analysis shows that with 3 faces, each 19cm by 19cm or so, you can fit 8 parallel trenches with a width of 1.25cm, leaving 9cm of unfilled area.

This means you can fit 19*8*3, or 456cm of trench into the interior 3 faces of the furnace.

The requirement is 420cm

Due to the fact that you need to lead wire outside of the device, and there are curves that need to be followed (8 curves, 4 per side facing the opening of the furnace), this theoretical trench length is reduced somewhat. I believe that 15cm loss is reasonable for this, so you should be left with close to 420cm of routing area.

Now I expect the wire to be kinda hokey to coil right, so I expect 10% losses in length possibly, but it should work out.


[Edited on 14-5-2018 by coppercone]

coppercone - 14-5-2018 at 11:19

4.5kW for a 7inch cube does sound a bit toasty.

It should result in quick heating times though.

The question now becomes:

What kind of temperature rate of rise is acceptable for the furnace firebricks?

I intend to use the dense kind found at lowes rated for 3000F for the inner shell. They are rather heavy.

https://www.homedepot.com/p/Rutland-4-5-in-x-9-in-x-1-25-in-...

But lowes brand. They weigh about 4lbs each.

I wanted durable brick that needs to be cut with a water cooled saw to avoid the problems I see with soft insulation brick eroding away. I intend to use the soft insulation bricks behind the main durable lining.

It is also my hope that the high conductivity of a liner shell will reduce how much heat is concentrated in the trenches dug out for the coils, so that there is less stress there during a fast start up.


On the other hand, there is the option of using this:

https://www.itccoatings.com/new-products/itc-100-ht-ceramic-...

Which will increase the reflectivity of the furnace interior.

Will it degrade life of coils?

how much will the kanthal wire sublime in a vacuum? I was thinking about going to vacuum furnace construction.

[Edited on 15-5-2018 by coppercone]

coppercone - 23-5-2018 at 08:24

I think a good idea for this would be to desifn a special door that can be used to manufacture white phosphorus or possibly vacuum distill metals.

I think you can wrap a pipe with kanthal to heat it like a short path lenght distillation aparatus

coppercone - 23-5-2018 at 10:38

My idea is to make a second door that has a fairly large square opening, then manufacture a bunch of small insinuative blocks, so that you can put the reactor inside of the furnace, then use your insulation blocks like bricks to make a rough wall around the reactor inlets/outlets

It won't contain all the heat, but if you make the mini-bricks fairly small then you can seal around it pretty good. I imagine them like 5mmx5mmxthickness of furnace wall. Afterwords you can use disposable sheet metal with cut outs to finish the seal so that the little insulation blocks don't come falling out in case the furnace is bumped accidentally or something and hold it in place with strong magnets or screws.. with silicone and possibly some hydroforming to make a domed/expanded disposable plate, which can fit some kaowool, that might hopefully make up for the non-ideal seal formed by the temporary brickwork.

THis way you can make custom reactors fairly quickly with a high degree of control and safety, compared to shit like making phosphine vapors in a tin can connected to some gas pipe lol

I can imagine having a shelf full of different door designs to be quite annoying if you are very interested in high temperature custom synthesis.

[Edited on 23-5-2018 by coppercone]

zed - 26-5-2018 at 15:38

Well, perhaps I was wrong, it is a question of perspective.

Maybe there is a very expensive super-furnace, that is the standard of the industry.

We just don't usually think of it as a furnace. It has another name.

https://www.youtube.com/watch?v=7i20brSAOLc

The Parr Pressure Reactor. Usually used at lower temperatures, a Tungten reaction vessel, might enable very high temperatures indeed.

Note: Construction is modular.



[Edited on 26-5-2018 by zed]

[Edited on 26-5-2018 by zed]

coppercone - 26-5-2018 at 17:07

axt made something like that before