# Capture cross-section issues using Shielding in G4.11.4

**URL:** <https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082>\
**Category:** Physics Processes, Models and Cross Sections\
**Created:** [April 14, 2026, 3:17pm UTC](https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082 "2026-04-14T15:17:48Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![Eric\_Esch](https://geant4-forum.web.cern.ch/user_avatar/geant4-forum.web.cern.ch/eric_esch/32/10774_2.png) [@Eric\_Esch](https://geant4-forum.web.cern.ch/u/Eric_Esch)\
**Post date:** [April 14, 2026, 3:17pm UTC](https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082/1 "2026-04-14T15:17:49Z")

</div>

Since the update to G4.11.4 the capture cross-sections `NeutronHPCaptureXS` seem to have issues to be loaded with the use of pre-defined physicslists like `Shielding` or `QGSP_BIC_HP`. This can be easily reproduced using the Hadr04 example:

1. Change `run01.mac` (in Hadr04) by commenting the line `/testhadr/phys/thermalScattering true` (this is for consistency, as Shielding does not have that option)

2. Change the Material to `/testhadr/det/setMat G4_Gd`. (Here it is the easiest to see, but from my testing also `G4_STAINLESS-STEEL` and `G4_lAr` seems to be effected).

3. Add following macros to only inspect nCapture (after `/run/initialize`): `/process/inactivate hadElastic neutron`, `/process/inactivate neutronInelastic neutron`

4. Change the energy to 0.1 eV (or lower) and increase number of primaries to 20 000 (or more).

5. comment the verbose commands (`/control/verbose 2`, `/run/verbose 2`) to not get blasted.

Now run it once with the default NeutronHP physics in this example. You will see the expected list of generated particles: Almost all captures on Gd-155 and Gd-157 (creating Gd-156 and Gd-158).

Now in `Hadr04.cc` replace `PhysicsList* phys = new PhysicsList;` with `G4VModularPhysicsList* phys = new Shielding;` (and add the include). Re-make and re-run the simulation. You should see completely wrong isotope production.

I can also quickly swap between Geant4 versions. If i go to Geant4.11.3.2 and use `Shielding` i see something different than the “standard” physics in Hadr04 but comparable. If i compare the two versions the issue occurs in PhysicsLists using `NeutronHPCaptureXS`, but not in `QGSP_BERT_HP` which uses `neutronCaptureHP`.

I did a lot of investigating to find the reason for this but i have no idea. I did see that in the release notes it says: “New hadronic datasets G4PhotonEvaporation-6.1.2, G4INCL-1.3, G4PARTICLEXS-4.2 are required and should be used.” but “G4PARTICLEXS-4.2” does exist in the data folder. I tried going through the source code on Github, as there were plenty of Changes for G4.11.4 but i did not really find anything immediately obvious (at least to me).

TLDR: Am i making a mistake or are pre-defined physicslists unusable for neutron physics in G4.11.4?

_Geant4 Version: geant4-11-04 [MT]_  
_Operating System: Ubuntu 20.04_  
_Compiler/Version:_ g++ 13.3.0  
_CMake Version: 3.28.3_

* * *

---

<div class="post-metadata">

**Author:** ![pico](https://geant4-forum.web.cern.ch/letter_avatar_proxy/v4/letter/p/46a35a/32.png) [@pico](https://geant4-forum.web.cern.ch/u/pico)\
**Post date:** [April 15, 2026, 7:47am UTC](https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082/2 "2026-04-15T07:47:25Z")

</div>

Hello @Eric_Esch ,

welcome to the Geant4 forum and thank you for your nice reproducible report.

Our colleague @dkonst, had a look at your issue and it seems there is an issue when combining physics that are element-driven and the ones that are more isotope-driven. We’ll back to you as soon as we have had a closer look at this.

Cheers,

/Pico

---

<div class="post-metadata">

**Author:** ![Eric\_Esch](https://geant4-forum.web.cern.ch/user_avatar/geant4-forum.web.cern.ch/eric_esch/32/10774_2.png) [@Eric\_Esch](https://geant4-forum.web.cern.ch/u/Eric_Esch)\
**Post date:** [April 15, 2026, 11:38am UTC](https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082/3 "2026-04-15T11:38:26Z")

</div>

Hello @pico,

Thanks for the reply, keep me updated!

Best,

Eric

---

<div class="post-metadata">

**Author:** ![Eric\_Esch](https://geant4-forum.web.cern.ch/user_avatar/geant4-forum.web.cern.ch/eric_esch/32/10774_2.png) [@Eric\_Esch](https://geant4-forum.web.cern.ch/u/Eric_Esch)\
**Post date:** [June 22, 2026, 2:12pm UTC](https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082/4 "2026-06-22T14:12:09Z")

</div>

Hey @pico ,

Are there any news regarding this topic? Especially because conda-forge is pushing out G4 11.4.2 it would be great to know the current status.

Best,  
Eric

---

<div class="post-metadata">

**Author:** ![pico](https://geant4-forum.web.cern.ch/letter_avatar_proxy/v4/letter/p/46a35a/32.png) [@pico](https://geant4-forum.web.cern.ch/u/pico)\
**Post date:** [June 22, 2026, 2:51pm UTC](https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082/5 "2026-06-22T14:51:58Z")

</div>

Hello,

Nothing to report yet. There was a period of heavy teaching and now it is summer. Stay tuned !

/Pico

---

<div class="post-metadata">

**Author:** ![pico](https://geant4-forum.web.cern.ch/letter_avatar_proxy/v4/letter/p/46a35a/32.png) [@pico](https://geant4-forum.web.cern.ch/u/pico)\
**Post date:** [November 18, 2026, 12:29pm UTC](https://geant4-forum.web.cern.ch/t/capture-cross-section-issues-using-shielding-in-g4-11-4/15082/6 "2026-11-18T12:29:15Z")

</div>

Hello,  
I have checked the [Geant4 PR #119](https://github.com/Geant4/geant4/pull/119) which is titled **“G4NeutronHPCaptureData: sample the capture target isotope by its own cross section.”** Its description identifies the same isotope-sampling bug, the same `Shielding` and `QGSP_BIC_HP` impact.. So **this PR should have fixed and and it will be available in the next release**

/Pico
