Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Nichrome Resistance Wire Consistently Failing

  1. Jan 10, 2015 #1
    Greetings, folks!

    I have spent the past six years developing a matching pair of 13"-square heating plates to mount to the inside bottom (front and rear) of a Sun-Mar Excel NE composting toilet drum. The heating plates are necessary to keep the compost in the composting drum between 70º and 100ºF.

    I have been using Chromel-A nickel chromium resistance wire (see http://www.amazonsupply.com/dp/B000FMW8X0/ref=sp_dp_g2c_asin) mounted to aluminum plates and it works very well! The challenge is that after an average of about eighteen months, the nichrome wire breaks at some point in the circuit, thus requiring replacing both heating plates. (If one has failed, prudence requires that both be replaced.)

    In the latest incarnation of each of the two assemblies, I used fifteen feet of 16-gauge nichrome wire, carefully laid out in a grid pattern, pressed between two anodized, 1/8" thick aluminum plates, bolted together with stainless steel bolts, lock nuts, and neoprene rubber washers. Nine bolts are used for each heating plate assembly to ensure that there is no point where the resistance wire is not in direct contact with both of the two plates.

    The finished plates are sealed shut with a large bead of 100% silicone sealant, inside the edges of the two plates. The resistance wires are connected to 12-gauge THHN solid-core wire at the point where they leave the heating plate assembly, and the splices are sealed and secured to the plate assembly with epoxy. A PID controller is used to maintain the temperature of the compost.

    It really does work quite well! Each assembly has a resistance of about 3.7 ohms and draws about 2.7A at a nominal 12-15V DC, fed by Trojan T-105 batteries, charged by solar panels. (Details HERE.)

    I am offering all this (perhaps excessive!) information because, once again, after seventeen months, one of the heating plate assemblies has stopped working. I knew as soon as it happened, because its reed-switch-controlled LED indicator light ( thank you! ) went out, and after careful inspection, the cause can only be that the resistance wire has, once again (for the fourth time!), broken inside the heating plate assembly.

    The two plates were tested extensively before mounting them inside the composting drum; it is a multi-day process to replace them, after the replacements have been assembled, and I am a monk, camped out in the wilderness, in the mountains of Utah. This is a really bad time for one of the heating plates to fail!!

    Can anyone offer any reason why such nichrome resistance wire, especially 16-gauge!, would fail so consistently, when it is in a sealed environment, and only gets its two heating plates up to no more than 140ºF, when tested in the summer in open-air conditions.

    Is there any better solution for heating these two heating plate assemblies? Any heat source outside of the composting drum is not possible, as it does get “a bit chilly” high up in the mountains of Utah! :-)

    Blessings and thanks in advance!,

    Richard Fairbanks
     
    Last edited by a moderator: May 7, 2017
  2. jcsd
  3. Jan 10, 2015 #2

    phinds

    User Avatar
    Gold Member
    2016 Award

    That doesn't sound like a real-world test to me since the use is in an enclosed space. Do you know what it gets up to in actual use? Even so, though, that doesn't sound like anywhere near high enough temp to cause it to break.

    Does the wire break at a bend or just somewhere along the length?
     
  4. Jan 10, 2015 #3

    Bystander

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    It does sound like a kinked hot spot. The care described in assembly argues against such a thing not being noticed. Next time you assemble a plate, you might try smearing heat sink compound between the anodized plates, or at least along the heating element to guarantee heat transfer away from any hot spots. Never had any trouble with a "run-away" high resistance spot burn through at those temperatures with similar assemblies in grad school.
     
  5. Jan 10, 2015 #4

    Doug Huffman

    User Avatar
    Gold Member

    Hot spots and corrosion were my guesses. Aluminum is pretty reactive, forms an oxide layer that is probably not a good HX surface.
     
  6. Jan 10, 2015 #5

    Bystander

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    He said, "anodized" which actually isn't that bad, but might abrade kinks in the filament wire aggravating the hot spot problem.
     
  7. Jan 10, 2015 #6

    meBigGuy

    User Avatar
    Gold Member

    In the absense of any other explanations, more data is needed. I would try periodically measuring the current during operation. If the current decreases after a month or two, it might indicate that a high resistance spot is developing. At that point you can disassemble and try to determine what is happening.

    Also, looking at the used plate that is still working might give you a clue. Is it heating unevenly. Has its resistance changed? Do you have access to a thermal camera or IR thermometer?

    Just rambling.
     
  8. Jan 10, 2015 #7

    nsaspook

    User Avatar
    Science Advisor

  9. Jan 10, 2015 #8

    Bystander

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    What're you trying to do? Break up a "helpful" speculation party with a bunch of ugly facts?
     
  10. Jan 10, 2015 #9

    NascentOxygen

    User Avatar

    Staff: Mentor

    Can you examine the break (both sides) under a low-power microscope? Looking for heat discoloration, narrowing, pitting, corrosion products.

    The bare nichrome is in direct contact with the anodized layer on the aluminium? What is the appearance of the aluminium in the spot where the wire break developed?
     
  11. Jan 11, 2015 #10

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    One wonders if there's a short to aluminum plate shortly before failure.
    First short doesn't hurt anything, but the second one completes a circuit.

    Could you use a Cal-Rod heater and slightly thicker plates to make up for the wider spacing ?

    Everyday RTV Silicone releases acetic acid during cure.

    http://www.logwell.com/tech/servtips/RTV.html
    Which type are you using?


    Can you post a picture of a failed plate's heated side with wires still in place ?
     
    Last edited: Jan 11, 2015
  12. Jan 11, 2015 #11

    Bystander

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    And let's give that man a cigar --- not something to have sealed between plates with a heating element.
     
  13. Jan 11, 2015 #12

    RonL

    User Avatar
    Gold Member

    Taking lessons from electric kiln design, might suggest a need for coiled or spring tension to keep expansion and shrinkage under control ?:oldwink:
     
  14. Jan 11, 2015 #13

    The Electrician

    User Avatar
    Gold Member

  15. Jan 11, 2015 #14

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

  16. Jan 15, 2015 #15
    Thank you all so much for your advice; it is very much appreciated!

    I sincerely apologize for not being able to respond for the past few days, but no sunlight = no electricity = no on-line access.

    To start, http://www.LotusDome.com/Heating_Plate.jpg [Broken] is a photo of one of the heating plates (10" x 12"), opened up after the first set of four hours of testing. The pale-white residue around the edges and the bolt holes is the (clear) RTV silicone.

    Please allow me to address the suggestions in order of posts:

    from phinds:
    The open-air testing, in the summer, was specifically done to test the maximum temperature the plates could ever achieve. In use, they are buried under a few inches of compost, inside a drum that vents through a four-inch pipe straight up to open air, which, in the winter can get below -20°F.

    In fact, liquids seeping through the compost and draining into the evaporation chamber—only three inches below the drum and the two heating plates bolted to the inside bottom of it—will regularly freeze in the winter, and if too deep will freeze to the bottom of the drum, while the heating plates are on!, making it unable to rotate. Regulating moisture content of the compost is critical!! ;-)

    The PID controller (using a teflon-encased temperature sensor set into the compost, alongside another temperature sensor for comparison) is programmed to keep the temperature of the compost at 80°F and due to the elevation, the heating plates are on almost every night of the year. For six months or more, they are on all day as well and even still, the compost can drop below 60°F during chilly nights or high winds.

    Thus, as per the above (hopefully not too excessive!) details, it doesn’t make sense that temperature could cause the wire to break after, what appears to be, a consistent amount of time.

    The breaks have, so far, all been in the middle of straight runs. I do not know where the latest break is, as I really need some well-considered course of action before I take a few days to completely disassemble and rebuild my only toilet, high on a mountain, in the middle of winter! The past breaks have appeared to be melts, sometimes with very small balls of copper forming at one or both ends of the break. That is why I went to great effort to ensure that the 16-gauge wire was pressed tightly between the two anodized plates.

    from Bystander:
    Thank you for the suggestion! Again, that is why I carefully pressed the wire between the two plates, but it certainly would help!

    Each of the anodized aluminum plates in this latest version (repeatedly tested to ensure no conductivity between the wire and the plate), has a layer of clear packing tape taped to it to also ensure no shorts between the wire and the plate. During the hours of 140°F testing (that could never be reached during actual use), the packing tape didn’t even get singed, much less melt. The first test actually used plain scotch tape (as shown in the above-linked photo, after testing) to ensure the runs across the plate remained evenly spaced. Both the initial scotch tape and the final packing tape were completely unfazed by the heat generated by the wire. Even if the packing tape had melted, which is exceedingly unlikely, the wire would not have shorted out to the plate.

    I was quite careful in unspooling the wire to ensure there were no kinks. Also, given that the plates are bolted together and maintain a steady range of (relatively) low temperatures (50-80°F) abrasion seems quite unlikely. But that’s just my opinion!

    from: meBigGuy
    Thank you for the suggestions! After a few months, I stopped measuring current draw at the drum because any wide changes would be noticed by monitoring at my TM-2020 meter. The current of the still-functional plate does not appear to have changed since it was made. (Please forgive me for not having more elaborate testing equipment!)

    from: Jim Hardy:
    A pair of shorts would result in the length of resistance wire being reduced and thus would result in significant changes in both temperatures and current. I saw no such results. (But that doesn’t mean it might not have suddenly occurred, long after being assembled, and right before breaking!)

    I have been using DAP 100% Silicone Caulking; I must assume it is an RTV silicone. I have used the same in the past to secure 20-gauge resistance wire to aluminum sheets, to avoid shorts. I never saw any degradation of the wire in any of the earlier assemblies, though they also broke after about 18 months.

    Could the RTV silicone possibly be the cause of 16-gauge wire consistently breaking/melting after 17-18 months of contact? Practically speaking, how likely is that, given that the heating plate assemblies have consistently failed, at relatively low temperatures?

    from: The Electrician:
    Thanks for the link! When I was refining the design, I tried to find such heating element films with no luck, thus, my home-made solution. They probably aren’t cheap, but they may be necessary!

    One critical issue that haunts me: Is there some consistent degradation over time of the nichrome resistance wire itself, simply due to current flowing through it, regardless of the temperature generated or any RTV silicone? The wire in each heating plate assembly is carefully protected from any liquids in the compost, and being pressed between two plates was specifically done to avoid any temperature fluctuations within the wire that might result from the wire pulling away from a plate.

    Thank you all very much, yet again, for any insights you might be willing to offer!,

    Richard Fairbanks
     
    Last edited by a moderator: May 7, 2017
  17. Jan 15, 2015 #16

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    You are certainly a thorough fellow !

    Do you know what type of Nichrome you have?

    Nichrome_Resistance_Wire_9155_1.jpg

    One possibility to rule out is chloride stress corrosion,
    It makes microcracks that grow along grain boundaries surprisingly fast
    especially under tensile stress.

    go to the 3M site and see if your tape is made from PolyvinylChloride

    and inquire of wire supplier about Chloride Stress Corrosion resistance for the alloy you have
    the ones with significant Fe are likely susceptible

    Search engines are toying with me tonight. Can't find a good paper on chloride stress corrosion in Nichrome. Too many different Nichromes, i guess..

    http://www.astm.org/SUBSCRIPTION/DIGITAL_LIBRARY/STP/PAGES/STP39166S.htm

    might be as simple as a different wire or tape.

    See bottom of page 9 here , the table showing chloride stress corrosion inception times for a few different alloys.
    http://www.haynesintl.com/pdf/h3179.pdf [Broken]
    their engineering department would likely suggest an alloy ....
     
    Last edited by a moderator: May 7, 2017
  18. Jan 15, 2015 #17

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

  19. Jan 16, 2015 #18

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    I have only just come to this thread but the first thing that struck me is that ordinary Electrical Immersion Heaters work fine for years and years - in hot water tanks. Why not use a simple immersion heater in a small s/s water tank and use that as an intermediate heating element? That way, someone else has taken care of the mechanical / electrical problems of the electrical heater and all you need to do is make a heat exchanger. The heater could even be remote from the composter and you could use convection to get the heat inside it. Car Coolant would allow you to operate right up to 100C without boiling / bursting.
     
  20. Jan 16, 2015 #19
    Hello Lotus -

    Note on the silicone - in addition to the noted acidic curing issue - most RTV Silicones are Hydroscopic - meaning they absorb water - add to this thermal cycling and you accelerate the effect.... so my first guess is you are getting the nichrome wet (or just moist) with an acidic solution.... not so good.

    You may want to look into a marine store - for a true water proof and harsh environment type sealant. -- Found this http://www.dxengineering.com/parts/umi-82180 [Broken]

    Looking at your plates - is is possible to get someone with a router to make channels for the wire? It looks like when you clamp the plates together - they are clamping against the wire - while good for thermal transfer - not so good for the wire - remember as this heats and cools it is moving back and forth- since it is very unlikely (impossible) the plates are clamped uniformly across all the wire there will be a couple locations that take all of the mechanical stress. ( with the above permatex - you may want to use a bead along all of the wire - to set the wire in place and probably have a better thermal conductivity than the air around the wire.

    I know you want to keep this as a DIY - and an electrical heater seems to be a simple enough system - making them work reliably in a real world application is really not trivial- and the compost is actually a pretty harsh environment.. So -- if the heaters are critical - you may want to consider some silicone strip heaters - you can get them for 12 V

    Side note : You seem to be doing excellent considering what you have been through and where you are! -- keep in touch.
     
    Last edited by a moderator: May 7, 2017
  21. Jan 16, 2015 #20

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    @Lotus
    I am surprised that you can source Resistance Wire as cheaply as a commercial heater, actually (the way the market works in these matters). Seriously, though, you want this to be a long-term working system, presumably, so it would be in your interest to treat it as a serious Engineering Project. That implies using available technology when it is not much more expensive than DIY. Engineering tends only to be 'fun' when the results work and survive for a reasonable time. I have to agree that it would be good to make it all for yourself, from scratch, but don't you also want a working composter?
    At the very least, I would suggest isolating the wires from the 'grot' with an intermediate layer of corrosion inhibiting coolant, in a tank.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Nichrome Resistance Wire Consistently Failing
  1. Nichrome wire troubles (Replies: 4)

  2. Nichrome Wire Heater (Replies: 2)

Loading...