HPT transient thermography

Why transient IR on HPT hardware

Steady thermography on cooled turbine vanes conflates inlet boundary-layer state, coolant redistribution, and metal conduction into a single snapshot. Transient techniques, typically a step change in mainstream or coolant conditions with time-resolved surface temperature capture, separate the external convection response from slow conduction effects when the test matrix is designed accordingly.

Leading-edge holing changes both aerodynamic loss and coolant paths. A pristine-blade baseline and a deteriorated hardware line therefore require distinct transient reductions, not a single emissivity map applied across rainbow sets. The turbomachinery deterioration thematic index and the companion aerodynamic and thermal manuscripts track how integrated metrics move with damage; this note documents the measurement chain those arguments depend on.

Data-reduction workflow

The sequence below reflects ECAT campaign practice. Facility-specific DAQ and rig hardware are logged on the project page; the documentation structure transfers to other engine-scale rigs.

1. Define the transient event: document whether the step is in mainstream total temperature, coolant flow, or both; record coolant-to-mainstream pressure ratio (CMPR) set points and dwell times
2. Calibrate radiometry: emissivity maps or spot checks on representative surface patches; note oxidation state and viewing angle per vane set
3. Align spatial and temporal grids: pixel pitch, integration time, and frame rate relative to expected thermal diffusion length scales
4. Reduce to observables: area-averaged wall temperature rise, inferred heat-transfer coefficient, or effectiveness definitions stated before comparison to CFD
5. Check Biot consistency: compare solid conduction time scales to external convection; use dimensional analysis and air properties for nondimensional groups
6. Archive raw and reduced data: link reduction scripts and calibration files from the project record alongside simulation cases

Biot number and conjugate interpretation

The Biot number compares internal conduction resistance to external convection resistance. When $Bi$ is not small, a lumped-capacitance assumption fails (a semi-infinite solid assumption fails for a different reason, once the heat penetration depth $\sqrt{\alpha_s t}$ in the solid is no longer small compared with the wall thickness) and inferred surface heat flux from a temperature trace alone can mislead both maintenance models and RANS wall models.

Here $h$ is the convective heat-transfer coefficient inferred from the transient reduction, $L_c$ a characteristic conduction length in the solid, and $k_s$ the effective solid conductivity including coating or oxidation layers where relevant. Simulation comparisons must state which wall boundary condition was applied (fixed temperature, fixed flux, or conjugate) and whether the experiment’s $Bi$ regime supports that choice. The general measurement posture, similarity, traceability, facility transfer, is developed in experimental aerothermal measurement.

Uncertainty budget (practical)

Separate bias sources that shift inferred $h$ or effectiveness from precision sources that widen confidence bands across repeated runs.

• Emissivity bias: oxidation, viewing angle, and spectral bandwidth; often dominates absolute temperature error
• Spatial resolution: pixel size relative to film-cooling holes and leading-edge features; area averaging must match CFD extraction windows
• Temporal sampling: frame rate vs. thermal response time; aliasing appears as false early-time slopes
• Background and reflections: rig hardware and neighbouring vanes; document subtraction and shielding per campaign
• Coolant metering: CMPR uncertainty propagates into effectiveness definitions when coolant flow is an input to the reduction

Pneumatic inlet surveys that set mainstream conditions carry their own bias budgets, see the 3-hole probe uncertainty note. Combined statements belong in validation reports, not in figure captions alone.

Pairing with simulation

RANS wall temperatures that match a thermography image while coolant splits differ by several percent are a common silent failure mode. Before tuning turbulence constants, verify that coolant-to-mainstream mass-flow ratio, inlet profile, and wall thermal boundary condition match the transient reduction inputs, the checklist in CFD-experiment validation is written for exactly this coupling.

For deteriorated geometry, simulation must represent hole blockage and leading-edge damage explicitly. A pristine-blade case cannot validate a rainbow set with oxidation holes without geometry updates traced to metrology, structured light scans and rig logs on the HPT project record document that path.