|
|
发表于 2004-2-9 13:07:37
|
显示全部楼层
[精彩]tinge等人关于流体数值模拟的问答录 (无内容)
[这个贴子最后由windsnow在 2004/02/10 00:52am 第 2 次编辑]
Flowermoon and tinge, I think you are both right about the relationships between modelers and models. Generally speaking, there are two different types of numerical models: models mainly used for scientific research to study a SPECIFIC physical process - mechanistic models, such as a DNS (Direct Numerical Simulation) model of a homogeneous fluid to study transition and development of turbulence (say, Type-A models); models mainly used for industrial/commercial applications that have to include ALL physical (and chemical) processes, such as a model with or without combustion processes under any geometric settings (say, Type-B models). When I made that statement about relationship between modelers and models (flowermoon called it "distance produces beauty") I was more or less thinking of Type-B models, which are the ones most people in this forum are interested in. Often, a single model of Type-B will need a big group of people to develop and will also need continuous refinement. On the one hand, Type-B commercial models like to include as many physical processes/options as possible. On the other hand, to make models workable one cannot include all the known physical processes in their precise forms, let alone the unknown or little known physical processes. As a result (of many people working on one model and many simply parameterized physical processes), the actual reliability of such a model may not be as good/trusted as a Type-A model that tinge most likely was thinking about. Generally speaking, a Type-A mechanistic model needs only one person or two people to develop (similar to the case that a child needs only a couple of two people to produce). Under such a circumstance, the relationship between the modeler and the model is more or less a parent-child relationship, which tinge also called it "beauty comes from zero-distance". The parent (modeler) not only knows each and every part of her/his child (the model) but she/he is also able to know the exact physics behind it because the only unknown part is the one she/he is going to study.
Although both flowermoon and tinge are right on their understanding to the modeler-model relationship, I think tinge should get a slightly higher marker than flowermoon at the end of their discussions because the word "modeler" rather than "modelers" was repeatedly used in their discussions. In other words, although I used "modelers" in my statement about the relationship, which corresponds to Type-B models, their discussions can be considered being focused on the Type-A models, which should have a "beauty comes from zero-distance" relationship.
On the surface, it appears that flowermoon plays a major role in the current Q-A (Question-and-Answer) discussions between tinge and flowermoon. It also appears that flowermoon has given good answers to all the questions raised by tinge. However, I believe that it is tinge who has been leading all the discussions! It seems that flowermoon knows quite a lot about numerical modeling of fluid but I think it is impossible for flowermoon to produce a good answer without a good question from tinge. I think tinge is artfully testing flowermoon's academic aptitude by slowly raising the level of difficulty of the questions. Although flowermoon might be right at guessing that tinge is a female, flowermoon has probably underestimated tinge's academic aptitude. I strongly believe that when all the keys to the puzzles are revealed at the end (if there is a happy ending) of the game it is most likely that tinge is much more experienced than flowermoon in numerical modeling of fluid.
I wish you both good luck and wish the discussions to continue for a long, long time.
|
|
IP卡
狗仔卡
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
显身卡
楼主
发表于 2004-2-8 18:27:56
