Dynamic load testing of roller screw ema's

Solo disponible en BuenasTareas
  • Páginas : 16 (3752 palabras )
  • Descarga(s) : 0
  • Publicado : 21 de agosto de 2012
Leer documento completo
Vista previa del texto

Dale E. Schinstock
The University oif Alabama
Box 870276
Tuscaloosa, AL 35487-0276
Phone: (205)348-1648 Fax: (205)348-6419

Tim A. H askew
The University olf Alabama
Box 870286
Tuscaloosa, AL 35487-0286
Phone: (205)348-1766

comparison with the ball screw. However, there is still some
question about their suitability in adverseenvironments and in
severe loading applications where hydraulics have typically
been employed. Since EMA designs involving roller screws
have a single load path through the screw, it is important to
qualify the types of loading that roller screws can withstand.
Analysis of mechanical transmission failure modes has been
identified as a key research activity necessary to bring electric
actuationinto h l l competition with hydraulics (Working Group
# 3, 1994). This is especially important for mission critical
actuators, even where multiple actuators are used. It is dificult
to design an EMA where a failed roller screw can be disengaged
from the load to allow redundant actuators to take over.
In the use of an EMA for applications such as thrust
vector control, large dynamic loadsmay be generated with a
small amount of motion of the actuator. This loading condition
is more severe t a large loads with large motion. In order for
lubrication to work effectively a significant amount of motion is
required to draw the lubrication into the contact area


In the Electromechanical Actuators (EMA) Laboratory
at The University of Alabama a dynamic load test standhas been
designed and built. This test stand uses large load, high
bandwidth, hydraulic actuation to generate load profiles under
force control. The test stand can accommodate EMA's up to six
feet in length. It can generate dynamic loads of up to 100,000 lb
at fundamental frequencies of up to 12 Hz against a stiff
environment. This test stand has been used to generate severe
loadingconditions on a large roller screw in an attempt qualify
the effects of large, high frequency loads on roller screw.
During the tests performed in the EMA Laboratory the
screw was fixed at one end and axial loads were applied to the
roller nut at the other end. Since the end opposite the nut was
fixed, only a small amount of relative rotation between the nut
and screw was achieved. This rotationwas the result of elastic
deformation (wind up) of the screw along the length between the
fixed end and the nut. This simulates a severe, but likely,
application of the roller screw.
The results of the tests performed demonstrate that
roller screws may be damaged by dynamic loading with load
magnitudes that are well within the static load rating of the
screw. While the damage that wasobserved is not catastrophic,
it would be expected to substantially decrease the life of the


The general configuration of the test stand in the EMA
Laboratory is shown in Figure 1. This test stand was designed to
produce large dynamic loads on a linear actuator, under force
control or position control. It was also designed to be
reconfigurable so thatdifferent actuators could easily be
mounted in the stand for testing. The test stand will accept
either rigidly mounted linear actuators, as shown in the figure, or
self contained actuators with pivoting end connectors, like clevis
mounts. Nominally, all components in the test stand were
designed to withstand at least 100 kip in the extension of the

Roller screws have beenproposed and used as the
mechanical transmission for many electromechanical actuators
(EMA's). In comparisonwith ball screws, they are well suited to
those applications where the loads are large and/or have a
significant shock load content. This is due to the increased
diameter of and number of contact surfaces of the roller screw in

hydraulic cylinder and 5 0 kip in the retraction of the...
tracking img