Larson-Miller / Hollomon-Jaffe tempering parameter
Tempering, stress relief and post-weld heat treatment are all diffusion-controlled processes, and diffusion depends exponentially on absolute temperature but only linearly on time. That is why a modest temperature increase buys far more than a long extension of soak time. The Larson-Miller parameter — used in the same algebraic form as the Hollomon-Jaffe tempering parameter — captures the trade-off in a single number: LMP = T(C + log₁₀ t) / 1000, with T the absolute temperature in kelvin and t the holding time in hours.
Two heat treatments with the same parameter should produce approximately the same degree of tempering, the same drop in hardness and the same relief of residual stress. This lets you answer the question that comes up constantly on site: a vessel was soaked at 620 °C for 2 hours instead of the specified 650 °C for a shorter period — is that acceptable? Convert both to a parameter and compare. It also underlies the standard trade-off tables in ASME VIII-1 UCS-56, which permit a lower PWHT temperature if the holding time is increased.
The constant C is material dependent. C = 20 is the conventional value for carbon and low-alloy steel tempering and is the default here; some Cr-Mo creep work uses 19.5 or a fitted value. Using the same C for both sides of a comparison is what matters; the absolute parameter value only means something against a curve for the same material and the same C.
Two limits are worth remembering. The parameter has nothing to say about the heating and cooling rates, which codes control separately to avoid distortion and re-hardening, and it says nothing about staying below the lower critical temperature Ac1. Exceeding Ac1 — roughly 720 °C for plain carbon steel and lower for some alloys — re-austenitises the steel and the whole basis of the equivalence collapses. Equivalence calculations are a tool for evaluating what happened, not a licence to substitute an unqualified cycle for the one in the procedure.
T[K] = T[°C] + 273.15 LMP = T[K] × (C + log₁₀ t) / 1000 t in hours, C typically 20 Equivalent time at T₂: t₂ = 10^( LMP × 1000 / T₂[K] − C )
- The parameter compares tempering effect only. It says nothing about heating and cooling rates, which codes control separately.
- Never exceed the lower critical temperature Ac1 - the equivalence is meaningless once the steel re-austenitises.
- Code minimum holding times (for example ASME VIII-1 UCS-56, typically 1 hour per 25 mm with a 15 minute minimum for P-No. 1) apply regardless of the calculated equivalence.
- Use the same constant C on both sides of any comparison. Absolute LMP values are only meaningful against a curve for the same material and the same C.
Reference: Larson and Miller, Trans. ASME 74, 1952; Hollomon and Jaffe, Trans. AIME 162, 1945. Code holding times per ASME BPVC Section VIII Div 1 UCS-56


