DNS 1-2 Quantification of Resolution: Difference between revisions

From KBwiki
Jump to navigation Jump to search
(New article page for DNS_1-2_quantification)
 
Line 7: Line 7:
= Quantification of resolution =
= Quantification of resolution =
==Mesh resolution==
==Mesh resolution==
Provide wall resolution in wall coordinates, both normally ("y+") and tangentially ("x+", "z+").
The grid resolution corresponds to uniform <math>\Delta x^+=12.2,\ \Delta z^+=6.3</math> and ranges from <math>\Delta y^+=0.16</math> near the wall (first solution point) to <math>\Delta y^+=4.6</math> at the centre of the channel. The near-wall Kolmogorov length scale, <math>\eta=\left(\bar{\rho}\nu^3/\epsilon\right)^{1/4}</math>, in wall units is <math>\eta^+=1.54</math>.
<!--Provide wall resolution in wall coordinates, both normally ("y+") and tangentially ("x+", "z+").
Evaluate typical turbulence length scales (Taylor microscale, Kolmogorov) and compare to local
Evaluate typical turbulence length scales (Taylor microscale, Kolmogorov) and compare to local
resolution. In case the case presents homogeneous directions, one could also provide spatial
resolution. In case the case presents homogeneous directions, one could also provide spatial
correlations between the velocity components. If possible provide computed temporal spectra at
correlations between the velocity components. If possible provide computed temporal spectra at
selected locations and relate to spatial resolution e.g. by using Taylor's hypothesis.
selected locations and relate to spatial resolution e.g. by using Taylor's hypothesis.-->
==Solution verification==
==Solution verification==
The solution accurately captures the near wall and log-law behaviour of averaged streamwise velocity profile as shown in [[lib:DNS_1-2_quantification_#figure2|Fig. 2]].
<div id="figure2"></div>
{|align="center"
|[[Image:Average_u_logscale.png|400px]]
|-
|'''Figure 2:''' Average streamwise velocity, <math>Re_\tau=180</math>,  vs data from Moser et al. (1999)                           
|}
The turbulent stress profiles agree well with the data from Moser et al.
<div id="figure3"></div>
{|align="center"
|[[Image:Channel_turbulent_stress.png|400px]]
|-
|'''Figure 3:''' Average streamwise velocity, <math>Re_\tau=180</math>,  vs data from Moser et al. (1999)                           
|}
The expected near wall behaviour of turbulent stresses are <math>\widetilde{u^{\prime\prime2}},\widetilde{w^{\prime\prime2}}\sim y^2,\ \widetilde{v^{\prime\prime2}}\sim y^4</math> and <math>\widetilde{u^\prime v^\prime}\sim y^3</math>. These stresses near the wall from the current set of simulations are shown below.
<gallery widths=300px heights=240 mode="packed-hover">
Image:Near_wall_uu.png|<math>\widetilde{u^{\prime\prime} u^{\prime\prime}}</math>
Image:Near_wall_vv.png|<math>\widetilde{v^{\prime\prime} v^{\prime\prime}}</math>
Image:Near_wall_ww.png|<math>\widetilde{w^{\prime\prime} w^{\prime\prime}}</math>
Image:Near_wall_uv.png|<math>\widetilde{u^{\prime\prime} v^{\prime\prime}}</math>
</gallery>
One way to verify that the DNS are properly resolved is to examine the residuals of the Reynolds-
One way to verify that the DNS are properly resolved is to examine the residuals of the Reynolds-
stress budget equations. These residuals are among the statistical volume data to be provided as
stress budget equations. These residuals are among the statistical volume data to be provided as

Revision as of 11:59, 7 October 2021


Front Page

Description

Computational Details

Quantification of Resolution

Statistical Data

Instantaneous Data

Storage Format

Quantification of resolution

Mesh resolution

The grid resolution corresponds to uniform and ranges from near the wall (first solution point) to at the centre of the channel. The near-wall Kolmogorov length scale, , in wall units is .

Solution verification

The solution accurately captures the near wall and log-law behaviour of averaged streamwise velocity profile as shown in Fig. 2.

Average u logscale.png
Figure 2: Average streamwise velocity, , vs data from Moser et al. (1999)

The turbulent stress profiles agree well with the data from Moser et al.

Channel turbulent stress.png
Figure 3: Average streamwise velocity, , vs data from Moser et al. (1999)

The expected near wall behaviour of turbulent stresses are and . These stresses near the wall from the current set of simulations are shown below.

One way to verify that the DNS are properly resolved is to examine the residuals of the Reynolds- stress budget equations. These residuals are among the statistical volume data to be provided as described in Statistical Data section.



Contributed by: Lionel Agostini — Imperial College London

Front Page

Description

Computational Details

Quantification of Resolution

Statistical Data

Instantaneous Data

Storage Format


© copyright ERCOFTAC 2024