Add CSS surface tension validation

[svchb]

Purpose

Add a focused validation layer for the balanced CSS surface-tension operator on top of the current surface-tension stacks.

Changes

• fit the initial CSS capillary acceleration against the Young-Laplace pressure jump in 2D and 3D
• report fitted and virial surface tension, represented surface measure, residual, and net capillary force
• measure the instantaneous 2D mode-2 restoring stiffness against the Rayleigh frequency
• commit four-resolution Young-Laplace references in each dimension and three Rayleigh stiffness references
• add reduced deterministic CI checks and document the validation scope and tolerances

Stack

This validation leaf is built on an integration branch combining #1263, #1270, and #1276. Fork branches cannot be selected as bases for pull requests into the upstream repository, so this PR targets main and temporarily shows all preceding commits.

Review-only CSS validation diff: svchb/TrixiParticles.jlOpen@integration/surface-tension-all...validation/css-surface-tension

The integration branch also contains a small fix preventing an outer trixi_include override from rewriting the nested Akinci resolution study:
svchb/TrixiParticles.jlOpen@f6224f8...integration/surface-tension-all

Scope

This PR adds validation code, committed references, tests, and documentation only. The Rayleigh check measures instantaneous restoring stiffness; it does not introduce a long-running dynamic-drop benchmark.

Verification

• complete validation target: 85/85, including Akinci 42/42 and CSS 10/10
• surface-tension tests: 319/319
• IISPH tests: 156/156
• EDAC tests: 96/96
• WCSPH tests: 224/224
• observed Young-Laplace order: 2.652 in 2D and 1.250 in 3D
• selected Young-Laplace error: 4.854% in 2D and 0.913% in 3D
• Rayleigh frequency error: 4.296%, 3.997%, and 3.920%
• net capillary force remained at roundoff
• JuliaFormatter 2.1.1 and git diff --check passed