aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorGravatar Remi Lehe <remi.lehe@normalesup.org> 2023-09-20 00:25:55 -0700
committerGravatar GitHub <noreply@github.com> 2023-09-20 09:25:55 +0200
commit23a896d3c990eb4b7e86ebd8cb1fad723287e5c6 (patch)
treef454f9098e99fb8c94a524880af9c151a7226ae2
parenta0a3db1638ac94896924dc2a75b7e7e07072f9c8 (diff)
downloadWarpX-23a896d3c990eb4b7e86ebd8cb1fad723287e5c6.tar.gz
WarpX-23a896d3c990eb4b7e86ebd8cb1fad723287e5c6.tar.zst
WarpX-23a896d3c990eb4b7e86ebd8cb1fad723287e5c6.zip
Prevent NaNs in Coulomb collision module (#4304)
* Prevent NaNs in collisions between particles * Avoid another division by 0 t# Please enter the commit message for your changes. Lines starting
-rw-r--r--Source/Particles/Collision/BinaryCollision/Coulomb/UpdateMomentumPerezElastic.H343
1 files changed, 176 insertions, 167 deletions
diff --git a/Source/Particles/Collision/BinaryCollision/Coulomb/UpdateMomentumPerezElastic.H b/Source/Particles/Collision/BinaryCollision/Coulomb/UpdateMomentumPerezElastic.H
index 7c2432f6e..3101047e2 100644
--- a/Source/Particles/Collision/BinaryCollision/Coulomb/UpdateMomentumPerezElastic.H
+++ b/Source/Particles/Collision/BinaryCollision/Coulomb/UpdateMomentumPerezElastic.H
@@ -102,184 +102,193 @@ void UpdateMomentumPerezElastic (
T_PR const g1s = ( T_PR(1.0) - vcDv1*inv_c2 )*gc*g1;
T_PR const g2s = ( T_PR(1.0) - vcDv2*inv_c2 )*gc*g2;
- // Compute the Coulomb log lnLmd
- T_PR lnLmd;
- if ( L > T_PR(0.0) ) { lnLmd = L; }
- else
- {
- // Compute b0 according to eq (22) from Perez et al., Phys.Plasmas.19.083104 (2012)
- // Note: there is a typo in the equation, the last square is incorrect!
- // See the SMILEI documentation: https://smileipic.github.io/Smilei/Understand/collisions.html
- // and https://github.com/ECP-WarpX/WarpX/files/3799803/main.pdf from GitHub #429
- T_PR const b0 = amrex::Math::abs(q1*q2) * inv_c2 /
- (T_PR(4.0)*MathConst::pi*PhysConst::ep0) * gc/mass_g *
- ( m1*g1s*m2*g2s/(p1sm*p1sm*inv_c2) + T_PR(1.0) );
-
- // Compute the minimal impact parameter
- constexpr T_PR hbar_pi = static_cast<T_PR>(PhysConst::hbar*MathConst::pi);
- const T_PR bmin = amrex::max(hbar_pi/p1sm, b0);
-
- // Compute the Coulomb log lnLmd
- lnLmd = amrex::max( T_PR(2.0),
- T_PR(0.5)*std::log(T_PR(1.0)+lmdD*lmdD/(bmin*bmin)) );
- }
-
// Compute s
- const auto tts = m1*g1s*m2*g2s/(inv_c2*p1sm*p1sm) + T_PR(1.0);
- const auto tts2 = tts*tts;
- T_PR s = n1*n2/n12 * dt*lnLmd*q1*q1*q2*q2 /
- ( T_PR(4.0) * MathConst::pi * PhysConst::ep0 * PhysConst::ep0 *
- m1*g1*m2*g2/(inv_c2*inv_c2) ) * gc*p1sm/mass_g * tts2;
-
- // Compute s'
- const auto cbrt_n1 = std::cbrt(n1);
- const auto cbrt_n2 = std::cbrt(n2);
- const auto coeff = static_cast<T_PR>(
- std::pow(4.0*MathConst::pi/3.0,1.0/3.0));
- T_PR const vrel = mass_g*p1sm/(m1*g1s*m2*g2s*gc);
- T_PR const sp = coeff * n1*n2/n12 * dt * vrel * (m1+m2) /
- amrex::max( m1*cbrt_n1*cbrt_n1,
- m2*cbrt_n2*cbrt_n2);
-
- // Determine s
- s = amrex::min(s,sp);
-
- // Get random numbers
- T_PR r = amrex::Random(engine);
-
- // Compute scattering angle
- T_PR cosXs;
- T_PR sinXs;
- if ( s <= T_PR(0.1) )
- {
- while ( true )
+ T_PR s = 0;
+ if (p1sm > std::numeric_limits<T_PR>::min()) {
+
+ // s is non-zero (i.e. particles scatter) only if the relative
+ // motion between particles is not negligible (p1sm non-zero)
+
+ // Compute the Coulomb log lnLmd first
+ T_PR lnLmd;
+ if ( L > T_PR(0.0) ) { lnLmd = L; }
+ else
{
- cosXs = T_PR(1.0) + s * std::log(r);
- // Avoid the bug when r is too small such that cosXs < -1
- if ( cosXs >= T_PR(-1.0) ) { break; }
- r = amrex::Random(engine);
+ // Compute b0 according to eq (22) from Perez et al., Phys.Plasmas.19.083104 (2012)
+ // Note: there is a typo in the equation, the last square is incorrect!
+ // See the SMILEI documentation: https://smileipic.github.io/Smilei/Understand/collisions.html
+ // and https://github.com/ECP-WarpX/WarpX/files/3799803/main.pdf from GitHub #429
+ T_PR const b0 = amrex::Math::abs(q1*q2) * inv_c2 /
+ (T_PR(4.0)*MathConst::pi*PhysConst::ep0) * gc/mass_g *
+ ( m1*g1s*m2*g2s/(p1sm*p1sm*inv_c2) + T_PR(1.0) );
+
+ // Compute the minimal impact parameter
+ constexpr T_PR hbar_pi = static_cast<T_PR>(PhysConst::hbar*MathConst::pi);
+ const T_PR bmin = amrex::max(hbar_pi/p1sm, b0);
+
+ // Compute the Coulomb log lnLmd
+ lnLmd = amrex::max( T_PR(2.0),
+ T_PR(0.5)*std::log(T_PR(1.0)+lmdD*lmdD/(bmin*bmin)) );
}
- }
- else if ( s > T_PR(0.1) && s <= T_PR(3.0) )
- {
- T_PR const Ainv = static_cast<T_PR>(
- 0.0056958 + 0.9560202*s - 0.508139*s*s +
- 0.47913906*s*s*s - 0.12788975*s*s*s*s + 0.02389567*s*s*s*s*s);
- cosXs = Ainv * std::log( std::exp(T_PR(-1.0)/Ainv) +
- T_PR(2.0) * r * std::sinh(T_PR(1.0)/Ainv) );
- }
- else if ( s > T_PR(3.0) && s <= T_PR(6.0) )
- {
- T_PR const A = T_PR(3.0) * std::exp(-s);
- cosXs = T_PR(1.0)/A * std::log( std::exp(-A) +
- T_PR(2.0) * r * std::sinh(A) );
- }
- else
- {
- cosXs = T_PR(2.0) * r - T_PR(1.0);
- }
- sinXs = std::sqrt(T_PR(1.0) - cosXs*cosXs);
-
- // Get random azimuthal angle
- T_PR const phis = amrex::Random(engine) * T_PR(2.0) * MathConst::pi;
- T_PR const cosphis = std::cos(phis);
- T_PR const sinphis = std::sin(phis);
-
- // Compute post-collision momenta pfs in COM
- T_PR p1fsx;
- T_PR p1fsy;
- T_PR p1fsz;
- // p1sp is the p1s perpendicular
- T_PR p1sp = std::sqrt( p1sx*p1sx + p1sy*p1sy );
- // Make sure p1sp is not almost zero
- if ( p1sp > std::numeric_limits<T_PR>::min() )
- {
- p1fsx = ( p1sx*p1sz/p1sp ) * sinXs*cosphis +
- ( p1sy*p1sm/p1sp ) * sinXs*sinphis +
- ( p1sx ) * cosXs;
- p1fsy = ( p1sy*p1sz/p1sp ) * sinXs*cosphis +
- (-p1sx*p1sm/p1sp ) * sinXs*sinphis +
- ( p1sy ) * cosXs;
- p1fsz = (-p1sp ) * sinXs*cosphis +
- ( T_PR(0.0) ) * sinXs*sinphis +
- ( p1sz ) * cosXs;
- // Note a negative sign is different from
- // Eq. (12) in Perez's paper,
- // but they are the same due to the random nature of phis.
- }
- else
- {
- // If the previous p1sp is almost zero
- // x->y y->z z->x
- // This set is equivalent to the one in Nanbu's paper
- p1sp = std::sqrt( p1sy*p1sy + p1sz*p1sz );
- p1fsy = ( p1sy*p1sx/p1sp ) * sinXs*cosphis +
- ( p1sz*p1sm/p1sp ) * sinXs*sinphis +
- ( p1sy ) * cosXs;
- p1fsz = ( p1sz*p1sx/p1sp ) * sinXs*cosphis +
- (-p1sy*p1sm/p1sp ) * sinXs*sinphis +
- ( p1sz ) * cosXs;
- p1fsx = (-p1sp ) * sinXs*cosphis +
- ( T_PR(0.0) ) * sinXs*sinphis +
- ( p1sx ) * cosXs;
- }
- T_PR const p2fsx = -p1fsx;
- T_PR const p2fsy = -p1fsy;
- T_PR const p2fsz = -p1fsz;
+ // Compute s
+ const auto tts = m1*g1s*m2*g2s/(inv_c2*p1sm*p1sm) + T_PR(1.0);
+ const auto tts2 = tts*tts;
+ s = n1*n2/n12 * dt*lnLmd*q1*q1*q2*q2 /
+ ( T_PR(4.0) * MathConst::pi * PhysConst::ep0 * PhysConst::ep0 *
+ m1*g1*m2*g2/(inv_c2*inv_c2) ) * gc*p1sm/mass_g * tts2;
- // Transform from COM to lab frame
- T_PR p1fx; T_PR p2fx;
- T_PR p1fy; T_PR p2fy;
- T_PR p1fz; T_PR p2fz;
- if ( vcms > std::numeric_limits<T_PR>::min() )
- {
- T_PR const vcDp1fs = vcx*p1fsx + vcy*p1fsy + vcz*p1fsz;
- T_PR const vcDp2fs = vcx*p2fsx + vcy*p2fsy + vcz*p2fsz;
- /* factor = (gc-1.0)/vcms; Rewrite to avoid subtraction losing precision when gc is close to 1 */
- T_PR const factor = gc*gc*inv_c2/(gc+T_PR(1.0));
- T_PR const factor1 = factor*vcDp1fs + m1*g1s*gc;
- T_PR const factor2 = factor*vcDp2fs + m2*g2s*gc;
- p1fx = p1fsx + vcx * factor1;
- p1fy = p1fsy + vcy * factor1;
- p1fz = p1fsz + vcz * factor1;
- p2fx = p2fsx + vcx * factor2;
- p2fy = p2fsy + vcy * factor2;
- p2fz = p2fsz + vcz * factor2;
- }
- else // If vcms = 0, don't do Lorentz-transform.
- {
- p1fx = p1fsx;
- p1fy = p1fsy;
- p1fz = p1fsz;
- p2fx = p2fsx;
- p2fy = p2fsy;
- p2fz = p2fsz;
- }
+ // Compute s'
+ const auto cbrt_n1 = std::cbrt(n1);
+ const auto cbrt_n2 = std::cbrt(n2);
+ const auto coeff = static_cast<T_PR>(
+ std::pow(4.0*MathConst::pi/3.0,1.0/3.0));
+ T_PR const vrel = mass_g*p1sm/(m1*g1s*m2*g2s*gc);
+ T_PR const sp = coeff * n1*n2/n12 * dt * vrel * (m1+m2) /
+ amrex::max( m1*cbrt_n1*cbrt_n1,
+ m2*cbrt_n2*cbrt_n2);
- // Rejection method
- r = amrex::Random(engine);
- if ( w2 > r*amrex::max(w1, w2) )
- {
- u1x = p1fx / m1;
- u1y = p1fy / m1;
- u1z = p1fz / m1;
-#ifndef AMREX_USE_DPCPP
- AMREX_ASSERT(!std::isnan(u1x+u1y+u1z+u2x+u2y+u2z));
- AMREX_ASSERT(!std::isinf(u1x+u1y+u1z+u2x+u2y+u2z));
-#endif
+ // Determine s
+ s = amrex::min(s,sp);
}
- r = amrex::Random(engine);
- if ( w1 > r*amrex::max(w1, w2) )
- {
- u2x = p2fx / m2;
- u2y = p2fy / m2;
- u2z = p2fz / m2;
+
+ // Only modify momenta if is s is non-zero
+ if (s > std::numeric_limits<T_PR>::min()) {
+
+ // Get random numbers
+ T_PR r = amrex::Random(engine);
+
+ // Compute scattering angle
+ T_PR cosXs;
+ T_PR sinXs;
+ if ( s <= T_PR(0.1) )
+ {
+ while ( true )
+ {
+ cosXs = T_PR(1.0) + s * std::log(r);
+ // Avoid the bug when r is too small such that cosXs < -1
+ if ( cosXs >= T_PR(-1.0) ) { break; }
+ r = amrex::Random(engine);
+ }
+ }
+ else if ( s > T_PR(0.1) && s <= T_PR(3.0) )
+ {
+ T_PR const Ainv = static_cast<T_PR>(
+ 0.0056958 + 0.9560202*s - 0.508139*s*s +
+ 0.47913906*s*s*s - 0.12788975*s*s*s*s + 0.02389567*s*s*s*s*s);
+ cosXs = Ainv * std::log( std::exp(T_PR(-1.0)/Ainv) +
+ T_PR(2.0) * r * std::sinh(T_PR(1.0)/Ainv) );
+ }
+ else if ( s > T_PR(3.0) && s <= T_PR(6.0) )
+ {
+ T_PR const A = T_PR(3.0) * std::exp(-s);
+ cosXs = T_PR(1.0)/A * std::log( std::exp(-A) +
+ T_PR(2.0) * r * std::sinh(A) );
+ }
+ else
+ {
+ cosXs = T_PR(2.0) * r - T_PR(1.0);
+ }
+ sinXs = std::sqrt(T_PR(1.0) - cosXs*cosXs);
+
+ // Get random azimuthal angle
+ T_PR const phis = amrex::Random(engine) * T_PR(2.0) * MathConst::pi;
+ T_PR const cosphis = std::cos(phis);
+ T_PR const sinphis = std::sin(phis);
+
+ // Compute post-collision momenta pfs in COM
+ T_PR p1fsx;
+ T_PR p1fsy;
+ T_PR p1fsz;
+ // p1sp is the p1s perpendicular
+ T_PR p1sp = std::sqrt( p1sx*p1sx + p1sy*p1sy );
+ // Make sure p1sp is not almost zero
+ if ( p1sp > std::numeric_limits<T_PR>::min() )
+ {
+ p1fsx = ( p1sx*p1sz/p1sp ) * sinXs*cosphis +
+ ( p1sy*p1sm/p1sp ) * sinXs*sinphis +
+ ( p1sx ) * cosXs;
+ p1fsy = ( p1sy*p1sz/p1sp ) * sinXs*cosphis +
+ (-p1sx*p1sm/p1sp ) * sinXs*sinphis +
+ ( p1sy ) * cosXs;
+ p1fsz = (-p1sp ) * sinXs*cosphis +
+ ( T_PR(0.0) ) * sinXs*sinphis +
+ ( p1sz ) * cosXs;
+ // Note a negative sign is different from
+ // Eq. (12) in Perez's paper,
+ // but they are the same due to the random nature of phis.
+ }
+ else
+ {
+ // If the previous p1sp is almost zero
+ // x->y y->z z->x
+ // This set is equivalent to the one in Nanbu's paper
+ p1sp = std::sqrt( p1sy*p1sy + p1sz*p1sz );
+ p1fsy = ( p1sy*p1sx/p1sp ) * sinXs*cosphis +
+ ( p1sz*p1sm/p1sp ) * sinXs*sinphis +
+ ( p1sy ) * cosXs;
+ p1fsz = ( p1sz*p1sx/p1sp ) * sinXs*cosphis +
+ (-p1sy*p1sm/p1sp ) * sinXs*sinphis +
+ ( p1sz ) * cosXs;
+ p1fsx = (-p1sp ) * sinXs*cosphis +
+ ( T_PR(0.0) ) * sinXs*sinphis +
+ ( p1sx ) * cosXs;
+ }
+
+ T_PR const p2fsx = -p1fsx;
+ T_PR const p2fsy = -p1fsy;
+ T_PR const p2fsz = -p1fsz;
+
+ // Transform from COM to lab frame
+ T_PR p1fx; T_PR p2fx;
+ T_PR p1fy; T_PR p2fy;
+ T_PR p1fz; T_PR p2fz;
+ if ( vcms > std::numeric_limits<T_PR>::min() )
+ {
+ T_PR const vcDp1fs = vcx*p1fsx + vcy*p1fsy + vcz*p1fsz;
+ T_PR const vcDp2fs = vcx*p2fsx + vcy*p2fsy + vcz*p2fsz;
+ /* factor = (gc-1.0)/vcms; Rewrite to avoid subtraction losing precision when gc is close to 1 */
+ T_PR const factor = gc*gc*inv_c2/(gc+T_PR(1.0));
+ T_PR const factor1 = factor*vcDp1fs + m1*g1s*gc;
+ T_PR const factor2 = factor*vcDp2fs + m2*g2s*gc;
+ p1fx = p1fsx + vcx * factor1;
+ p1fy = p1fsy + vcy * factor1;
+ p1fz = p1fsz + vcz * factor1;
+ p2fx = p2fsx + vcx * factor2;
+ p2fy = p2fsy + vcy * factor2;
+ p2fz = p2fsz + vcz * factor2;
+ }
+ else // If vcms = 0, don't do Lorentz-transform.
+ {
+ p1fx = p1fsx;
+ p1fy = p1fsy;
+ p1fz = p1fsz;
+ p2fx = p2fsx;
+ p2fy = p2fsy;
+ p2fz = p2fsz;
+ }
+
+ // Rejection method
+ r = amrex::Random(engine);
+ if ( w2 > r*amrex::max(w1, w2) )
+ {
+ u1x = p1fx / m1;
+ u1y = p1fy / m1;
+ u1z = p1fz / m1;
+ }
+ r = amrex::Random(engine);
+ if ( w1 > r*amrex::max(w1, w2) )
+ {
+ u2x = p2fx / m2;
+ u2y = p2fy / m2;
+ u2z = p2fz / m2;
+ }
#ifndef AMREX_USE_DPCPP
AMREX_ASSERT(!std::isnan(u1x+u1y+u1z+u2x+u2y+u2z));
AMREX_ASSERT(!std::isinf(u1x+u1y+u1z+u2x+u2y+u2z));
#endif
- }
+
+ } // if s > std::numeric_limits<T_PR>::min()
}