Résumé : While clusters of galaxies are considered as one of the most important cosmological probes, the standard spherical modelling of the dark matter and the intracluster medium is only a rough approximation. Indeed, it is well established both theoretically and observationally that galaxy clusters are much better approximated as triaxial objects. However, investigating the asphericity of galaxy clusters is still in its infancy.
I will review this topic which is currently gathering a growing interest from the cluster community. i will begin by introducing the triaxial geometry, then discuss the topic of deprojection and demonstrate the need for combining different probes of the cluster's potential, and finally review the different works that have been addressing these issues.
Then i will present a general parametric framework intended to simultaneously fit complementary data sets (X-ray, Sunyaev Zel'dovich and lensing data), and show an illustration on galaxy cluster Abell 1689.
I will show that, for strong lensing clusters, a triaxial model generally allows to lower the inferred value of the concentration parameter compared to a spherical analysis. This may alleviate tensions regarding, e.g., the over-concentration problem. However, i will stress that predictions from numerical simulations rely on a spherical analysis of triaxial halos.
Since triaxial analyses will have a growing importance in the observational side,  I advocate the need for simulations to be analysed in the very same way, allowing reliable and meaningful comparisons. Besides, methods and algorithms intended to derive the three dimensional shape of galaxy clusters should be extensively tested on simulated multi-wavelength observations in order to quantify the limitations of the triaxial model, which still represents an approximation  of what a real galaxy cluster might be.