Triple coronal Hard X-Ray source observed by STIX during a failed eruption of a filament.

Solar Hard X-Ray (HXR) emission is observed typically in the form of localized, rather compact sources. Majority of these sources are flare related, however there is growing evidence, observational and theoretical, that we can expect HXR emission from places not directly related to primary energy release regions. From this point of view, failed eruptions are phenomena which are very interesting. Failed eruptions are events which at an early phase develop like a typical eruption which evolves into CME. However, for unclear reasons these eruptions stop abruptly somewhere in the solar corona. Many mechanisms are suspected to be responsible for stopping an eruption i.e. kink instability leading to stable configuration of erupting flux tube, magnetic tension within erupting structure or interaction and reconnection with an overlying coronal magnetic field. The last one may lead to electron acceleration and production of high temperature regions emitting in the HXR. There are only a few observations, from past instruments, showing some evidence of HXR production in failed eruptions. Recently, the Solar Orbiter has been launched giving a completely new perspective for observation and analysis of coronal dynamic events. Apart from others, it carries onboard Extreme-Ultraviolet Imager (EUI) and Spectrometer/Telescope for Imaging X-rays (STIX) telescopes open an opportunity to understand the physics of failed eruptions. The new EUV and HXR observations are important for at least two reasons. First, SO can approach the Sun closer than 0.3 a.u. which increase spatial resolution and sensitivity of telescopes, giving ocasion for registering weak HXR sources which can not be observed with previous instruments. Second, SDO/AIA instrument operating on the Earth orbit can provide stereoscopic context for SO/EUI images which may help to investigate deeper the geometry of the eruption and causes of its braking. Here, we present a very well observed flare accompanied by a failed eruption. The event occurred when longitudinal separation of the Earth and SO was 17 degrees. From the Earth perspective the event was visible very close to the west limb of a solar disc. For SO it was a behind-the-limb event with occulted footpoints which gave us a very good view, especially in the HXR range, of emission coming from the solar corona. During a flare's impulsive phase, the eruption accelerated to the velocity of a few hundreds kilometers per second and after six minutes it stopped abruptly at the height of 100 000 km above the solar surface. The eruption reconnected with overlying coronal loops leading to occurrence of at least three regions observed by STIX. The lowest source seems to be a typical, flare related coronal source, possibly consisting of two spatially unresolved sources located close to the primary reconnection site. The other two sources are located higher in the corona where reconnection of the eruption with an overlying magnetic field was observed. These sources behave significantly differently than the lowest source. Namely, their evolution is more gradual and seems to be driven by direct heating not the evaporation of chromospheric plasma which is a case for the lowest source.