Automatic grid adaptation and error evaluation for unstructured finite volumes

Automatic grid reﬁnement Monogrid Error Estimate

Automatic

grid

adaptation

and error

Michel VISONNEAU

estimation

Fluid Mechanics Laboratory - CNRS UMR 6598 Ecole Centrale Nantes, Nantes, FRANCE

Accuracy and uncertainty in ﬂow simulation ONERA Symposium December 3, 2010

Michel VISONNEAU

Automatic grid adaptation and error estimation

Outline

of

the

Automatic grid reﬁnement Monogrid Error Estimate

presentation

Main motivations,

Automatic grid (de)reﬁnement, reﬁnement criteria,

2D monogrid error estimates,

Limitations and perspectives.

Michel VISONNEAU

Automatic grid adaptation and error estimation

Outline

of

the

Automatic grid reﬁnement Monogrid Error Estimate

presentation

Main motivations,

Automatic grid (de)reﬁnement, reﬁnement criteria,

2D monogrid error estimates,

Limitations and perspectives.

Michel VISONNEAU

Automatic grid adaptation and error estimation

Outline

of

the

Automatic grid reﬁnement Monogrid Error Estimate

presentation

Main motivations,

Automatic grid (de)reﬁnement, reﬁnement criteria,

2D monogrid error estimates,

Limitations and perspectives.

Michel VISONNEAU

Automatic grid adaptation and error estimation

Outline

of

the

Automatic grid reﬁnement Monogrid Error Estimate

presentation

Main motivations,

Automatic grid (de)reﬁnement, reﬁnement criteria,

2D monogrid error estimates,

Limitations and perspectives.

Michel VISONNEAU

Automatic grid adaptation and error estimation

Automatic grid reﬁnement Monogrid Error Estimate Main motivations

A non-exhaustive list of objectives To reduce the computational time by using more efﬁciently the available computational resources, To quantify and control the errors involved in a simulation To increase theaccuracyandefﬁciencyof numerical simulations, To obtain acceptable numerical errors near structures oflocalised ﬂow phenomena, To capture: sharpfree-surface,bilge or tip vortices,ﬂow behind propeller plane

To account for unsteady ﬂow where theposition of the local features changes in time,

All these objectives are fulﬁlled if one introducesgrid adaptivitybased on a reliableerror estimate.

Michel VISONNEAU Automatic grid adaptation and error estimation

Automatic grid reﬁnement Monogrid Error Estimate Main motivations

A non-exhaustive list of objectives To reduce the computational time by using more efﬁciently the available computational resources, To quantify and control the errors involved in a simulation To increase theaccuracyandefﬁciencyof numerical simulations, To obtain acceptable numerical errors near structures oflocalised ﬂow phenomena, To capture: sharpfree-surface,bilge or tip vortices,ﬂow behind propeller plane

To account for unsteady ﬂow where theposition of the local features changes in time,

All these objectives are fulﬁlled if one introducesgrid adaptivitybased on a reliableerror estimate.

Michel VISONNEAU Automatic grid adaptation and error estimation

Automatic grid reﬁnement Monogrid Error Estimate Main motivations

A non-exhaustive list of objectives To reduce the computational time by using more efﬁciently the available computational resources, To quantify and control the errors involved in a simulation To increase theaccuracyandefﬁciencyof numerical simulations, To obtain acceptable numerical errors near structures oflocalised ﬂow phenomena, To capture: sharpfree-surface,bilge or tip vortices,ﬂow behind propeller plane

To account for unsteady ﬂow where theposition of the local features changes in time,

All these objectives are fulﬁlled if one introducesgrid adaptivitybased on a reliableerror estimate.

Michel VISONNEAU Automatic grid adaptation and error estimation

Automatic grid reﬁnement Monogrid Error Estimate Main motivations

A non-exhaustive list of objectives To reduce the computational time by using more efﬁciently the available computational resources, To quantify and control the errors involved in a simulation To increase theaccuracyandefﬁciencyof numerical simulations, To obtain acceptable numerical errors near structures oflocalised ﬂow phenomena, To capture: sharpfree-surface,bilge or tip vortices,ﬂow behind propeller plane

To account for unsteady ﬂow where theposition of the local features changes in time,

All these objectives are fulﬁlled if one introducesgrid adaptivitybased on a reliableerror estimate.

Michel VISONNEAU Automatic grid adaptation and error estimation