diff --git a/doc/jcl_template.py b/doc/jcl_template.py
index 7c4f5eb..81bea31 100755
--- a/doc/jcl_template.py
+++ b/doc/jcl_template.py
@@ -335,10 +335,13 @@ def __init__(self):
                          'support': [0, 1, 2, 3, 4, 5],
                          # True or False, enables flutter check with k, ke or pk method
                          'flutter': False,
-                         # flutter parameters for k and ke method
+                         # Flutter parameters for k and ke method
                          'flutter_para': {'method': 'k', 'k_red': np.linspace(2.0, 0.001, 1000)},
-                         # flutter parameters for pk method
-                         # 'flutter_para': {'method': 'pk', 'Vtas': np.linspace(100.0, 500.0, 100)},
+                         # Flutter parameters for pk method
+                         # There are two implementations of the PK method: 'pk_schwochow', 'pk_rodden'
+                         # Available mode tracking algortihms: 'MAC', 'MACXP', 'MAC*PCC' (recommended)
+                         # 'flutter_para': {'method': 'pk', 'Vtas': np.linspace(100.0, 500.0, 100),
+                         #                  'tracking': 'MAC*PCC'},
                          },
                         ]
         # End
diff --git a/loadskernel/equations/frequency_domain.py b/loadskernel/equations/frequency_domain.py
index a742e74..695654e 100644
--- a/loadskernel/equations/frequency_domain.py
+++ b/loadskernel/equations/frequency_domain.py
@@ -628,11 +628,19 @@ def eval_equations(self):
 
     def calc_eigenvalues(self, A, eigenvalues_old, eigenvectors_old):
         eigenvalue, eigenvector = linalg.eig(A)
-        # To match the modes with the previous step, use a combination of modal assurance criterion and pole correlation.
-        # This improves the handling of complex conjugate poles.
-        MAC = fem_helper.calc_MAC(eigenvectors_old, eigenvector)
-        PCC = fem_helper.calc_PCC(eigenvalues_old, eigenvalue)
-        idx_pos = self.get_best_match(MAC * PCC)
+        # To match the modes with the previous step, use a correlation cirterion as specified in the JCL.
+        # The MAC matrix is used as default.
+        if 'tracking' not in self.simcase['flutter_para']:
+            MAC = fem_helper.calc_MAC(eigenvectors_old, eigenvector)
+        elif self.simcase['flutter_para']['tracking'] == 'MAC':
+            MAC = fem_helper.calc_MAC(eigenvectors_old, eigenvector)
+        elif self.simcase['flutter_para']['tracking'] == 'MACXP':
+            MAC = fem_helper.calc_MACXP(eigenvectors_old, eigenvector)
+        elif self.simcase['flutter_para']['tracking'] == 'MAC*PCC':
+            # This is a combination of modal assurance criterion and pole correlation and improves the handling of complex
+            # conjugate poles.
+            MAC = fem_helper.calc_MAC(eigenvectors_old, eigenvector) * fem_helper.calc_PCC(eigenvalues_old, eigenvalue)
+        idx_pos = self.get_best_match(MAC)
         eigenvalues = eigenvalue[idx_pos]
         eigenvectors = eigenvector[:, idx_pos]
         return eigenvalues, eigenvectors