Test & Validation

What we do

An analysis is worth what is known about how far it can be trusted. For us validation is not a separately sold service but part of every job — though sometimes it is needed on its own.

Scope

• Measurement correlation — comparing your existing test data against simulation, with the size and cause of any deviation
• Supplier test validation — whether a supplier's published curve holds under your operating conditions
• Third-party analysis review — an independent assessment of someone else's assumptions, mesh quality and results
• Uncertainty estimation — discretisation uncertainty by Richardson extrapolation, sensitivity studies

When there is no CAD

The part to be analysed may have no CAD data — old equipment, a supplier that never provided it, or an as-built that differs from the drawing. In that case we capture the geometry by laser scanning.

Both the turbine geometry of a hydroelectric plant and the low- and high-temperature centrifugal pumps used on a diesel engine were previously modelled this way and validated against test results.

If there is a deviation

We do not hide it. The size of the deviation, its likely cause and its effect on the conclusion are written plainly in the report. It is easy for an analysis to say it matched; it is valuable when it says where it did not.