Skip to main content
Electro SourceSiemens Software Partner — Platinum Expert
Customers · Anonymized engagements

Where simulation, measurement, and correlation meet a qualification date.

Three anonymized engagements structured as Challenge → Approach → Measured Result → Products Used. Figures shown in [__] brackets are placeholders until they are replaced with verified data from the underlying engagement — Electro Source does not publish numbers we have not measured.

  1. Case 01
    SiC power module supplier · Automotive Tier-1 qualification

    Closed a 20%+ Rth model gap and qualified against AQG 324 without an extra prototype build.

    Challenge
    Simulation and measured junction-to-case Rth diverged by more than 20% on a new SiC module. The TIM bond-line was not behaving as the datasheet assumed, and the Tier-1 customer was expecting AQG 324 v04.1 power-cycling evidence.
    Approach
    Electro Source ran ASTM D5470 characterization on the incumbent and candidate TIMs across the assembly's actual bond-line range with DynTIM, correlated the measured properties into the Flotherm module model with T3Ster structure-function data, and executed the AQG 324 profile on Power Tester with in-situ Rth monitoring.
    Measured result
    Model gap reduced from [__%] to under [__%] against measurement. AQG 324 evidence delivered in [__ weeks] on the first fixture, without an additional prototype cycle.
    Products used
    • DynTIM
    • T3Ster
    • Flotherm
    • Power Tester
  2. Case 02
    Solid-state lighting OEM · Outdoor architectural luminaire

    Identified TIM pump-out as the driver of a [__%] flux loss and requalified with a substitute material.

    Challenge
    A high-power LED luminaire lost more of its initial luminous flux than LM-80 / TM-21 modelling had predicted after [__] hours of field operation. Package degradation and board-level TIM pump-out were both plausible.
    Approach
    T3Ster structure-function measurements separated die-level Rth drift from board-level bond-line change. DynTIM benchmarked three candidate replacement TIMs across the assembly's pressure and thickness range. FLOEFD updated the luminaire thermal model with the measured substitute.
    Measured result
    Root cause isolated to board-level TIM pump-out. Substitute TIM met the luminaire's revised luminous-maintenance target of [__ hours] with margin.
    Products used
    • T3Ster
    • DynTIM
    • FLOEFD
  3. Case 03
    Hyperscale operator · AI compute rack thermal audit

    Evaluated single-phase vs two-phase immersion for a [__ kW] AI rack before committing to a coolant.

    Challenge
    An AI compute rack exceeded ASHRAE TC 9.9 A2 air-cooled limits under the target workload. The operator needed to evaluate single-phase dielectric immersion against two-phase before selecting a coolant partner and CDU vendor.
    Approach
    STAR-CCM+ resolved coupled fluid, thermal, and boiling behaviour for both single-phase and two-phase candidates. Flomaster sized the facility water loop for both cases. HEEDS drove the manifold-geometry study across [__] variants.
    Measured result
    Recommended architecture met the [__ °C] junction target with [__%] pump-power margin. Coolant and CDU vendor selection made on the basis of measured pressure-drop and dry-out margin, not vendor claims.
    Products used
    • STAR-CCM+
    • Flomaster
    • HEEDS
Bring your engagement

Every engagement is scoped, quantified, and delivered against a defined date.

Talk to a Thermal Specialist →