I am having to do a path length correction for a metal tube (stainless steel 316, 16ppm/C) from 0-100 degC. I need to correct to within approximately 1 ppm. How much do typical engineering metals deviate from linearity over this temperature range?

  • 2
    $\begingroup$ My usual path of finding these things out is: step 1) check the matweb page on the material in question (matweb is great). If not there, proceed to step two. Step 2) use the manufacturer links on the matweb page to check the technical specs the manufacturers have (Only once have I ever found that the manufacturers didn't have the data). If still not there, go to step 3. Step 3) Choose a reasonably similar alternate material and go back to step 1. If there are no reasonably similar materials, go to step 4. Step 4) Cry. $\endgroup$
    – Jim
    Commented Jan 30, 2015 at 15:59
  • $\begingroup$ Thanks - never come across matweb before. I'm mainly electronics and s/w, but have a physics background and am doing some physics R&D $\endgroup$
    – user56903
    Commented Jan 30, 2015 at 16:47

1 Answer 1


This page seems to give decent coverage of the topic, including this figure:enter image description here

They cite the empirically deduced thermal expansion of austenitic stainless steel to be: $$ \frac{\Delta l}{l} = 10^{-6}(16 + 4.76\times10^{-3}T - 1.243\times10^{-6}T^2)\cdot(T-20) $$ with $T$ having units of degC.

  • $\begingroup$ Thanks - seems there is a T^2 term in there that throws it off a little bit. $\endgroup$
    – user56903
    Commented Jan 30, 2015 at 12:27
  • $\begingroup$ Could you include the relevant information from the link in the body of your post (in case the target site is unreachable or goes permanently offline)? $\endgroup$
    – pentane
    Commented Jan 30, 2015 at 13:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.