Skip to main content Accessibility help
  • Print publication year: 2005
  • Online publication date: January 2010

Pr.3 - How good are measurements? Do they really confirm theory?



It is curious how people working in different disciplines seek and see validation in different kinds of experiments. Only the liquid helium up-the-walls measurement, described in the preceding subsection, seems to satisfy most people that the Lifshitz theory quantitatively accounts for measured forces (see note in the preceding subsection). As with the historical comments, this review of measurements is not intended to be exhaustive.

Force-balance measurements and atomic-beam measurements give qualitatively or semiquantitatively satisfying results, but they present worries about the smoothness of the surfaces and about the effective “zero” of separation at contact. For example, with no fitting parameters, the atomic-beam measurements already described give the predicted 1/l3 power law for the energy of interaction between atom and surface, but then there is a stubborn 60–75% overestimate of the coefficient computed by Lifshitz theory fed with good spectral data versus the experiments that admit little adjustment (see note 34 in the subsection titled “Measurably strong”). This level of disagreement might be resolved by better spectra or by better modeling of the surface structure. Surface roughness is often the fingered culprit, but specific attempts to formulate and compute its consequences fail. Formulations that include roughness are often based on pairwise summation and must themselves be validated by systematic measurement or by comparison with exact solutions.

That's the worst. Several kinds of measurements reassure us that the Lifshitz strategy of converting spectra into forces works reliably enough to capture the key features of van der Waals forces.