Hello,

I am using geant4.31 with the neutron_hp model to simulate low
energy neutron scattering in matter.

The angular distribution for elastic scattering shows a dependency on
the direction of the impinging neutron, which obviously is meaningless.

This problem can be traced back to the calculus of the momentum of the
neutron in the C. M. system.

I solved the problem by introducing the following lines of code in the
method G4ParticleChange * G4NeutronHPElasticFS::ApplyYourself()

//(replacing the line (line number 194):
//G4ThreeVector tempVector(en*sinth*cos(phi), en*sinth*sin(phi), en*cos(th
eta) );
//

//...
      G4ThreeVector np=theNeutron.GetMomentum();
      G4double vartheta=np.theta();
      G4double varphi=np.phi();
      G4ThreeVector tempVector;
      tempVector.setX(cos(theta)*sin(vartheta)*cos(varphi)
                      +sin(theta)*cos(phi)*cos(vartheta)*cos(varphi)
                      -sin(theta)*sin(phi)*sin(varphi)  );
      tempVector.setY(cos(theta)*sin(vartheta)*sin(varphi)
                      +sin(theta)*cos(phi)*cos(vartheta)*sin(varphi)
                      +sin(theta)*sin(phi)*cos(varphi)  );
      tempVector.setZ(cos(theta)*cos(vartheta)
                      -sin(theta)*cos(phi)*sin(vartheta)  );
      tempVector *= en;

      //G4ThreeVector tempVector(en*sinth*cos(phi), en*sinth*sin(phi), en*cos(th
eta) );


//...



I believe the effect of this change ought to be visible in the angular
distributions of neutrons whenever the initial direction
is not along the z axis, in each collision.

Thanks.

Cheers,
jose' pinto da cunha