International collaboration on the vacuum structure of full QCD using large-scale parallel computers

利用KEK的SR8000和RIKEN的SX 7进行的蒙特-卡罗模拟研究主要取得了以下结果：（1）有限温全QCD的DIK小组研究：国际合作DESY和ITEP小组。在16^3 × 8和24^3 × 10晶格上，用Nf=2的0（α）改进的三叶草费米子作用量确定了相变温度.利用前人在Nt=4和Mt=6时的数据，我们首次得到了手征极限和连续极限下的相变温度，尽管所用的夸克质量不够轻。我们正在使用KEK的新BlueGene机器继续模拟较轻的夸克质量。（2）约束机制的规范不变性：约束是一种物理现象，因此必须有一个规范不变性机制来解释约束。到目前为止，使用非定域规范的阿贝尔投影方法可以很好地解释禁闭的对偶Meissner效应，我们使用Riken的SX 7并进行了广泛的蒙特-卡罗模拟，发现即使在朗道规范中，阿贝尔对偶Meissner效应也很好地工作。然而在规范中，没有阿贝尔单极子，而是磁位移电流挤压电场。这种图像预计将在一般情况下工作，所以我们正在各种局部规范（称为坏规范）中进行模拟以检查预期。

The following results have been obtained mainly from Monte-Carlo simulation studies using SR8000 at KEK and SX7 at RIKEN :(1)DIK group study of finite-tenperature full QCD :International collaboration with DESY and ITEP groups. The phase transition temperature is determined on lattices of 16^3X8 and 24^3X10 using an 0(alpha) improved clover fermion action with Nf=2 light quarks. Using the previos data at Nt=4 and Mt=6,we have obtained for the first time the phase transition temperature in the chiral and the continuum limits, although the quark mass used is not light enough. We are continuing the simulations at lighter quark masses using a new BlueGene maschine at KEK.(2)Gauge invariance of the confinement mechanism :Confinement is a physical phenomenon and so there must be a gauge-invariant mechanism explaining confinement. Up to now abelian projection methods using some non-local gauges can explain the confinement beautifully as the dual Meissner effect.Using SX7 at Riken and performing an extensive Monte-Carlo simulations, we have found even in Landau gauge, the abelian dual Meissner effect works good. However in the gauge, there are no abelian monopoles, but instead the magnetic displacement currents squeezes the electric field.This kind of picture is expected to work in general and so we are doing simulations in various local gauges (known as a bad gauge) to check the expectation.