こちらで何回か質問をさせて頂いておりますzukkiと申します。
本解析はテストケースのため格子点数が少なく(40万点程度)しておりますが、200万点程度で行った別の解析でも同様の結果となりました。
CFD onlineや本グループにおいて、色々と情報を集め、解決方法を模索したのですが、未だ解決に至らなかったため、質問をさせて頂きました。
以下、解析の手順となります。
1. Gridgenにより回転領域と静止領域を作成し、Fluent形式で出力
2. Fluent形式からOF形式へメッシュの変換
3. 回転領域に対して、setSetで全領域を選択(boxToCell)し、setsToZones
4. mergeMeshesにより回転領域と静止領域を一体化
5. boundary, U, Pなど境界条件の設定
6. MRFSimpleFoam (10000stepまで計算)
boundary
RWall
{
type wall;
nFaces 11992;
startFace 996076;
}
IR1
{
type cyclicAMI;
nFaces 9832;
startFace 1008068;
matchTolerance 0.0001;
neighbourPatch IS1;
transform noOrdering;
}
RUPart
{
type patch;
nFaces 3360;
startFace 1017900;
}
IS1
{
type cyclicAMI;
nFaces 9832;
startFace 1021260;
matchTolerance 0.0001;
neighbourPatch IR1;
transform noOrdering;
}
SWall
{
type wall;
nFaces 12832;
startFace 1031092;
}
SUPart
{
type patch;
nFaces 2400;
startFace 1043924;
}
U
{
RWall
{
type fixedValue;
value uniform (0 0 0);
}
RUPart
{
type slip;
}
SWall
{
type fixedValue;
value uniform (0 0 0);
}
SUPart
{
type slip;
}
IR1
{
type cyclicAMI;
value $internalField;
}
IS1
{
type cyclicAMI;
value $internalField;
}
大変お忙しいところ誠に申し訳ありませんが、
もしよろしければ、皆様方のお知恵をお貸し頂ければ幸いです。
気になったことがあるため,投稿させて頂きます。
ohbuchiさんからの情報にあったサイトでは,nonRotatingPatches
にAMI面を指定するとの記述がありました。今回のケースでも,指定されていますか?
However, when using AMI with MRF, the AMI patches should be added to
the non-rotating patches in the constant/MRFZones dictionary, i.e. for
your case:
nonRotatingPatches (AMI1_1 AMI1_2 AMI2_1 AMI2_2);
この部分の指定ができておらず,AMI面にも回転の寄与が与えられていることはないのでしょうか?