Ball Screws

Feed Drive Positioning Error

Machine feed drives that use ball screws that incorporate rotary encoders are prone to position
measurement error as a result of thermally induced expansion because the ball screw drive serves
a double function: On the one hand it must be as rigid as possible to convert the rotary motion of
the servomotor to linear feed motion while, conversely, it must serve as a precision measuring
standard.
This twofold function forces a compromise because both the rigidity and the thermal expansion
depend on the preloading of the drive's ball nut and the fixed bearing, and both the axial rigidity
and the moment of friction are roughly proportional to the preloading.
Drive friction

The recirculating balls within the ball nut generate most of the friction within a typical ball screw
feed drive system. Besides rolling through their races and the adjoining ball screw threads, the
balls are subjected to micro-slip caused by the relative motion in compressed contact areas.
Additional friction occurs because the ball races do not completely hold the balls from all sides,
causing them to wobble much like tennis balls rolling down a gutter. The wobble produces
continual pressing and pushing with occasional slippages. The friction is further aggravated by
the absence of a retaining device in most ball nuts that can separate the balls and eliminate
high surface pressures.
In research performed by engineers at Heidenhain (www.heidenhain.com), measurements of the
moment of friction, various preload forces and rotational speeds of a typical ground precision
ball screw with a 40 mm diameter and 10 mm pitch confirmed the high share of solid-body
friction and mixed friction in ball screw drives at low speeds. Viscous friction dominated at high
speeds; however, for the ball screw described, normal machining feed rates were shown to lie far
below the speeds at which the moment of friction is at its minimum. The opposite was true at
rapid traverse feed rates.
The feed rates at which this ball screw was at optimum efficiency therefore seldom occurred.
The moment of friction was only slightly dependent on the axis load of the ball nut.
To increase rapid traverse velocity, either the pitch or the rotational speed of the ball screw must
increase. In the last 5 years, the maximum permissible speed of such ball screws has doubled;
however, due to the continually increasing requirement for acceleration, the preloading and
therefore the friction of the ball nut could not be reduced. Ball screw drives, therefore, generate
significantly more heat than before and will generate even more in the future.
According to Heidenhain research, if a feed drive ball nut has a typical preload of 3 kN and a
missing wiper, the unit's no-load or frictional moment would be 0.5 to 1 Nm. In rapid traverse at
a ball screw speed of 2000 rpm, approximately 100W to 200W of frictional heat would be
generated in the ball nut. Such an amount of heat is enough to cause expansion of the adjoining
feed components, and with typical time constants of 1 to 2 hours, the positioning error due to
thermal expansion can be in the magnitude of 0.1 mm, depending on the nature of the part
program. This positioning error therefore outweighs the thermally induced structural deformation
and geometric error of machining centers.
Ball screw bearing design
A suggested rule of thumb for thermal expansion is that, over the entire length of a cold ball
screw 1 meter in length, the ball screw grows by approx. 0.5 to 1 m after every double (back-
and-forth) stroke. This expansion accumulates within the time constant. After every new part
program, a feed drive system as just described would require approximately 1 hour to attain a
thermally stable condition. This also applies for interruptions in machining.
Differing types of heat expansion behavior depends on the types of bearing systems used to
support the ball screw. For example, a ball screw using fixed/floating bearings will expand freely
away from the fixed bearing in accordance with its temperature profile. The thermal zero point of
such a feed axis lies at the location of the fixed bearing. This means that theoretically no thermal
shift would be found if the ball nut is located at the fixed bearing. All other positions will be
affected by thermal expansion of the ball screw.
The situation is more complex in the case of fixed/fixed bearings. Ideally, rigid bearings would
prevent expansion of the ball screw at its end points. However, this would require considerable
force. To prevent expansion of a ball screw with 40 mm diameter, 2.6 kN of force must be
applied to the ball screw per degree Celsius of temperature increase.
A typical angular-contact ball bearing would quickly fail under any large increase in
temperature. Under real conditions, the rigidity of the purportedly fixed bearings with their seats
lies in the area of 800 N/m. This means that as the temperature of the ball screw increases, the
bearings deform significantly, and the end points of the ball screw move by 20 to 30 m in the
direction of the force generated by heat.
This means that by designing fixed bearings at both ends, the expansion could be halved. The
thermal zero point of the feed axis seems to lie at the midpoint of the traverse range. This is also
expected because the bearings have approximately equal rigidity and the ball screw was heated
evenly over its entire length.
A fixed/preloaded bearing design is often used as a sort of pressure valve. With a typical preload
of 50 m/m, one would expect that such a bearing configuration would behave like a fixed/fixed
combination up to a temperature increase of approx. 5 deg. K, and beyond that, like a
fixed/floating combination.
Countermeasures
Many ball screw systems use a rotary encoder attached to the screw's drive. The encoder counts
number of rotations, as well as identifies partial turns, for the control computer to convert into
position data. However, such an encoder cannot recognize if the overall length of the ball screw
has changed. If a feed system is subject to error due to thermal expansion, a linear encoder can
be added, or installed as a substitute for the rotary encoder. A linear encoder provides data based
on the actual position of the ball nut, and usually incorporates a thermally stable material in its
construction.
Another measure to use coolant conducting hollow ball screws. However, the circulation of the
coolant requires a hole in the ball screw and, for rotating ball screws, rotating bushings near the
screw bearings. Apart from the sealing problems, the hole reduces the ball screw's mechanical
rigidity in its already weak axial direction. The greatest problem, however, is a sufficiently
accurately temperature control of the coolant.
Research is underway on compensation of thermal deformation within the control software, with
the aid of analytic models, neural networks and empirical equations. However, the main focus of
these studies is in the deformation of the machine tool structure as a result of internal and
external sources of heat. There is little interest in investigation into compensation of axis drift.
Thus, as machine tool accuracy and velocity requirements increase, linear encoders for position
measurement should be taken into consideration when deciding on a feedback

Accuracy Grades

For precision ball screws, there are a number of standards which set forward basic specifications
for everything from materials to mathematical tolerances. These include the ANSI-B5.48, DIN
69051, ISO 3408, and JIS B1191 standards.
The most important specification for any ball screw is the lead error and lead accuracy. Lead
error is the difference between a specified distance and the actual travel distance of the lead
screw nut. This parameter, expressed as an error over a specified travel distance, helps determine
the accuracy rating of the ball screw.
Common types of ball screws include C-series (JIS standard) and Cp-series (equivalent to ISO)
for positioning, and Ct-series for transport. With respect to accuracy grades of ball screws, the
convention is that the lower the number the lower the error and thus the better the accuracy. So
for instance, an accuracy grade C0 precision ball screw is more accurate than a C2, a C3 is more
accurate than a C5, and so forth.
Precision ball screw lead accuracy conforms to four characteristics spelled out in the JIS
Standards; ep, vu, v300, and v2. These stand for:
ep : tolerance on specified travel: obtained by subtracting the specified travel from the actual
mean travel
vu : the maximum range relative to the effective length of the thread
v300 : the maximum range relative to the length of 300mm anywhere within the effective length
of thread
v2 : the maximum range which corresponds to any single rotation (2 rad) within the effective
length of thread