The CRYO group has a long experience in cryogenics for space applications. Indeed, since the late eighties, several space transportation applications requiring cryogenic fluids, especially nitrogen and oxygen (both liquid and gaseous) were studied.

10000image cryo

Test rigs available at the cryogenic test site of the CRYO group mainly concern quite delicate high rotational speed components of cryogenic rocket engines. That is to say bearings and dynamic seals. These rigs gave successfully important results for the understanding and the improvement of the running of the so called components.

The CRYO group started with the study of some HM7B cryogenic engine (viz. the engine of the upper stage of Ariane, the european launcher) features as: the gas generator firing sequence, pressure losses measurement into the injectors and liquid oxygen jet visualization into the gas generator. For the same engine, the CRYO group was also in charge to understand the oxygen turbopump ball bearing behavior and to validate modifications of this bearing. The validation tests had been performed by achieving around 300 tests. Cool-down tests in liquid oxygen atmosphere for the “chamber” and “purge” valves of the Vulcain engine LOx circuit were also performed at the same time (Vulcain is the main cryogenic engine of Ariane).

In parallel, tribological test campaigns on pin-on-disk tribometer in both LOx and LN2 were carried out. These experiments were conducted in the framework of new cryogenic valve developments. Moreover, the CRYO group payed particular attention to liquid oxygen compatibility. Among other things, these tests were dedicated to the understanding of tribo-oxidation, foil bearing applications and new cage materials for ball bearing.

Recently, the CRYO group has been chosen to study the new seal package design of the Vinci engine (Vinci is the new cryogenic engine of the upper stage of Ariane 6, the future european launcher). The project has necessitated the design and the manufacture of a new test bench.



The role of a ball bearing cage is to separate the balls and to equivalently distribute them around the whole bearing. However, under certain circumstances, the cage exhibits dynamical instabilities. Indeed, repetitive impacts of the cage with the rolling elements or the rings may lead to an erratic behavior. Those persistent rebonds within the bearing induce a significant increase in the bearing power losses and also in the frictional bearing torque. In the worst-case scenario this could even result in a ball bearing seizure.  

Several space applications are directly affected by the cage instabilities. As an example, this issue is a prime concern for the reaction wheels of satellites, which require the value of the bearing frictional torque to be correctly controlled. However, the value of this torque cannot be evaluated anymore if cage instability occurs. Some cage instability problems are also pointed out in cryogenic engines of launchers.

It appears from the specialized literature that the dynamical cage instabilities are still not well understood and consequently not well mastered. The experimental characterization of this phenomenon is extremely difficult because of the multiplicity of the parameters potentially involved (cage geometry, lubrication, bearing working conditions,...). Although some mathematical models exist to simulate the cage dynamics, they are not able to clearly identify the origin of the cage instabilities. Friction between the cage and its environment is often mentioned as the main trigger of such instabilities. Nevertheless, explanations concerning the influence of the friction on the dynamics of the cage remain ambiguous. As a result, no technical solution has ever been proposed to definitely fix the problem. By considering the case of the reaction wheel, one of the main potential artifices is to deliver a small amount of lubricant when an unstable phase is detected. However, this kind of solution does not systematically lead to successful results.

In the framework of a GSTP (General Support Technology Programme) with ESA (European Space Agency), a new mathematical model of the cage dynamics is currently developed. This model has to correctly represent the interactions of the cage with its environment (viz. the rings and the rolling elements of the bearing). Also, the model has to be easy to handle without necessitating extensive computational resources. This mathematical tool will be used to identify the key parameters of the ball bearing that lead to cage instabilities. Hence, technical solutions to the cage instability issue could be proposed.

Share this page