Page 1 of 6 123 ... LastLast
Results 1 to 10 of 59

Thread: annealing copper head gaskets

  1. #1
    Join Date
    Aug 2009
    Location
    rural eastern South Dakota
    Posts
    934

    Default annealing copper head gaskets

    I haven't found anyone who has annealed copper head gaskets, such as those with Indian Chiefs. A dozen or so individuals have tried to explain the process, and wikipedia explains that heating and quenching does not equate to hardening with copper. Must I consult the metallurgists from college?
    To what temperature must we HEAT the copper? What COLOR will evidence this heat? By what means, oxy-acet torch?
    Is air-cooling satisfactory? Or is a slow-cooling in vermiculite (or floor-dry) more effective? .. or quenching?
    Is there a simple "guage" to judge a copper gasket's hardness to determine the need to anneal?

    Is there a good reference beyond wiki (I would assume so)? .. if I can't find an experienced opinion...

  2. #2
    Join Date
    Jul 2012
    Posts
    371

    Default

    Quote Originally Posted by fillibuster View Post
    Must I consult the metallurgists from college?
    No, but that's what you're doing now...

    Quote Originally Posted by fillibuster View Post
    To what temperature must we HEAT the copper? What COLOR will evidence this heat? By what means, oxy-acet torch?
    Is air-cooling satisfactory? Or is a slow-cooling in vermiculite (or floor-dry) more effective? .. or quenching?
    Is there a simple "guage" to judge a copper gasket's hardness to determine the need to anneal?
    What happens with Cu isn't really different than with steel in that it is the size of the crystallites/grains that determine the hardness. A large Cu crystal is relatively soft, which means the Cu atoms can be pushed around fairly easily (note: fairly easy as compared with, say, iron). However, as this is done grains break up into smaller grains and the boundaries between grains (technically called, not surprisingly, "grain boundaries") have much higher resistance to motion than is the case within a grain/crystallite itself. The overall "hardness" of a piece of copper is a geometrical average of the hardness of the Cu within the grains and that of the grain boundaries, with an appropriate factor for the percentage of each making up the piece of copper.

    Refrigerator tubing is very soft because it was annealed at the factory so the crystallites in it are large and so there are relatively few grain boundaries. However, one only has to bend it a few times before it becomes quite hard because much of the mechanical energy required to bend the tubing goes into the creation of smaller grains by the formation of grain boundaries.

    OK, now to annealing. It turns out that heating a piece of copper to cherry red (~1400 oF) for a few seconds is sufficient to allow the Cu atoms near the grain boundaries to rearrange themselves, eliminating the boundaries, and resulting in the entire piece of Cu becoming soft again. Lower temperatures would work, but the time required decreases exponentially with increasing temperature so anything below cherry red would take too long for our purposes (note: it really is an exponential function of temperature, i.e. I haven't used this term in some loose fashion).

    Once formed, the large crystallites in the gasket retain their size irrespective of how fast the cooling takes place. This is different than the case of steel, but there the reason for the hardness isn't quite the same. With steel "second phases" such as carbides serve as the "pinning sites" to make it difficult for atoms to move (i.e. to make the steel hard) and they need time to form. As a result, how fast steels are cooled determine their final hardness. However, unlike steel, for copper there are no alloying elements (Mo, Co, etc.) included to determine the structural properties so only the grain size matters. As I said, the size of those grains (i.e. hardness of the Cu) does not depend on cooling rate. After heating to cherry red you can let it cool slowly in air (or even slower wrapped in a blanket), or drop it in a bucket of water, and the hardness will be the same.

    Another point is the entire gasket does not have to be cherry red at the same time. Acetylene is faster, but propane works. Dangle the gasket from a piece of wire and start at one end with the torch. Get whatever region cherry red that you can (a few square inches) and keep waving the torch around there for a few sec. Then slowly advance the torch and heat up the adjacent region. Think of your torch as painting the gasket red and with the paint drying behind the torch as it advances around the gasket. After you've advanced all the way around the gasket you can either allow it to cool in air, in which case you'll have to polish away quite a bit of black oxide that forms, or drop it in a bucket of water. Doing the latter will save you time.

    As for a gauge, it certainly is possible to measure the hardness of Cu with a Vickers, Brinell-type indenter. However, doing so only will tell you the hardness in one tiny location, whereas you need to know that the entire gasket is soft before you use it. Given how easy it is to anneal a gasket -- maybe 5 min. including time to fill the bucket of water -- I never measure the hardness even though I have the instruments to do it. I just fire up the torch.
    Last edited by BoschZEV; 04-06-2015 at 10:41 AM.

  3. #3
    Join Date
    Jan 2014
    Location
    Louisville, Kentucky, USA, Earth
    Posts
    35

    Default

    If I may clarify one little point from BoschZEV's excellent explanation. Work-hardening of a used copper gasket is a different type of hardening than the hardening that occurs when quenching a piece of steel. The fact that steel is hardened by the same process by which copper is annealed is a frequent source of confusion.

    Steel is hardened by manipulating the types of crystal structures which form at different temperatures and is specific to alloys of iron and carbon. (This BoschZEV's pinning sites. ) Work hardening has to do with grain size and occurs in most metals.

  4. #4
    Join Date
    Aug 2009
    Location
    rural eastern South Dakota
    Posts
    934

    Default

    You fellas don't know how pleased I am to be so rewarded. Ask and ye shall find, right here on the AMCA's version of wikipedia!
    I was all signed up for metallurgy (and I forget what else) when the wife told me she was pregnant again. BoschZEV, you have salvaged my loss of a semester! And fciron, you too are bringing me back to my materials studies. I left school to support the family, but the classes never left me.
    Please pardon me if I remain skeptical, not with your answers, but with the copper gaskets. They seemed to be fairly hard when new, stiff actually. Are they supposed to compress a little, to help fill in voids? Or are they famous for being re-useable? What are the convincing selling points?
    Again, thanks for your discussion, I think it's invaluable.

  5. #5
    Join Date
    Jul 2012
    Posts
    371

    Default

    Quote Originally Posted by fillibuster View Post
    They seemed to be fairly hard when new, stiff actually. Are they supposed to compress a little, to help fill in voids? Or are they famous for being re-useable?
    Copper gaskets are used on ultra-high vacuum (UHV) Conflat flanges to make the seal. The flange is designed with a shallow knife edge that displaces a small amount of the copper to form a seal that is entirely metallic (i.e. no organic grease). The Cu Conflat gaskets are roughly the same thickness as a head gasket. Straight out of the package they feel stiff, but Cu is a metal, after all, so you should expect it to be a lot stiffer than, say, a thin piece of plastic.

    Young's Modulus is the way stiffness is measured, and Cu comes in at 117 GPa. Wrought iron or steel is only ~2x that, while something like Teflon is only ~1%. So, in other words, a Cu head gasket should feel stiff. But under the pressure of head bolts it displaces easily enough to follow all the small waves and ripples of two surfaces that are not perfectly flat nor match perfectly to each other.

  6. #6
    Join Date
    Jul 2010
    Location
    Vermont
    Posts
    692

    Default

    Quote Originally Posted by fillibuster View Post
    You fellas don't know how pleased I am to be so rewarded.
    Me too! (means I was looking for something definitive, too!) Thanks for replying to the AMCA forum BoschZEV, FCIron, and Fillibuster...all good!
    Pisten Bulley is Harry Roberts in Vermont.

  7. #7
    Join Date
    Jul 2012
    Posts
    371

    Default

    Quote Originally Posted by fillibuster View Post
    Are they supposed to compress a little, to help fill in voids? Or are they famous for being re-useable? What are the convincing selling points?
    Oops, I forgot to address these questions. When used for a head gasket it's not so much that the Cu compresses (although it does a little), it's that it is soft so it follows the contours. Take a look at the following image I found on the web:



    Imagine trying to reproduce the Snap-On logo as best as possible by rubbing a stick back and forth across a piece of 0.002" shim stock that was made of 1) spring steel, or 2) soft copper. Clearly the Cu would deform to follow the contours of the logo much better than the spring steel. Although the contours on top of the cylinder, and on the bottom of the head, aren't nearly as pronounced as the logo on this wrench, deforming to follow those contours is what you need the head gasket to do, and annealed Cu is excellent for the task. It is soft enough to deform, yet strong enough to resist the pressure of combustion even on the hottest of days.

    As for being reusable, each time you anneal the Cu a little bit of the surface gets oxidized so you couldn't do it an infinite number of times. I've only needed to anneal a given head gasket a few times but I have little doubt I could do it at least 20 times before contamination might become an issue. Practically speaking, there's not much difference between 20 and infinite in this case (unless someone is a really bad engine builder, but even then they'll probably strip the head bolts or set fire to the garage before reaching 20 anneals...).

  8. #8
    Join Date
    Aug 2009
    Location
    rural eastern South Dakota
    Posts
    934

    Default

    qu: "deforming to follow those contours"
    I forgot about those contours ..... because I don't have any .... makes me a little nervous, hesitant. The "head shop" thought he found a "warp" in the top of the cylinders, and planed them off smooth. just a few thousandths. But I sense a disadvantage now.

  9. #9
    Join Date
    Jun 2001
    Location
    Central Illinois, USA
    Posts
    4,032

    Default

    Filibuster!

    You are talking about a Chief with cast-iron heads, right?

    ....Cotten
    AMCA #776
    Dumpster Diver's Motto: Seek,... and Ye Shall Find!

  10. #10
    Join Date
    Jul 2012
    Posts
    371

    Default

    Quote Originally Posted by fillibuster View Post
    qu: "deforming to follow those contours"
    I forgot about those contours ..... because I don't have any ....
    Unless your cylinder and head are flat and smooth enough to wring together, like a pair of gage blocks, you do have microscopic contours that a Cu gasket deals with. Further, even starting with flat surfaces, if you tighten down only one bolt a gap typically will appear on the other side (yes, I know we don't tighten one bolt at a time, but this illustrates the issue).

    The details depend on the shape of the head so no general answer can be given, but what is happening is when the bolt compresses the, say, 1" thickness of material under the bolt by 0.001" the ~1" of material in the vicinity expands by ~0.0003" (the amount of expansion perpendicular to the compression is given by Poisson's Ratio which typically is ~1/3). Since only the material under the bolt is compressed the horizontal expansion happens only to the bottom of the head and thus the mating surface is forced upwards (the same thing happens to the cylinder, doubling the separation).

    Simultaneously tightening opposing head bolts doesn't solve this. Although the circumference determined by the bolt circle pretty much keeps that annular region in good contact, the portion of the head in the center of the cylinder rises into the shape of a section of a sphere of very large radius. While this separation doesn't matter in the center of the cylinder, it does matter in the region between the bolt circle and the edge of the cylinder bore.

    The point this illustrates is clamping the two flat objects together necessarily results in distortion of the mating surface. An annealed Cu gasket deforms to follow the contours created by the distortion.
    Last edited by BoschZEV; 04-07-2015 at 09:14 AM.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •