Reducing Design Time for Linear Motion Systems
- Thursday, 01 December 2011
The life of the linear motion system can be predicted based
on its operational profile; simply, how many hours per day, days
per week, and weeks per year the ball screw will be run. For
more complex applications or more refined life prediction,
one needs to build a detailed, comprehensive motion profile
breaking down the movements to basically straight segments.
Each segment of the motion profile would require information
as to the speed at the beginning and end of the segment, the
time duration of the segment, and the torque during the segment.
Determine the positional accuracy and repeatability that your application requires. For example, inch ball screws are typically produced in two grades – Precision and Precision Plus. Precision grade ball screws would be used in applications requiring relatively coarse movement or those utilizing linear feedback for positional location. Precision Plus grade ball screws would be used when repeatable positioning within microns is critical and no linear feedback device is used. While Precision grade screws have greater cumulative variation over the useful length of the screw, Precision Plus grade screws limit accumulation of lead error, providing more precise positioning over the screw’s entire useful length.
Sizing and Selection
Charts provided by linear motion systems suppliers can be
a time-saving shortcut to proper sizing and selection of linear
motion systems. We’ll use a three-axis welding gantry
application as an example to demonstrate how to select and
size ball screws using catalog formulas (Figure 2). The ball
screw runs the entire length of the x-axis and is supported on
either end by bearing supports. For simplicity, we will define
the nut mounting as flanged, the material as alloy steel, the
thread direction as right-hand, and the product series as
metric. The system orientation in this application is horizontal,
with a screw-driven design and the length of the x-axis
being 6 meters. It will use fixed ends with a thermally stable
A load of 2,668.9 Newtons (600 lbs.) is applied by a carriage riding on profile rails. The travel length is 4.5 m (177.165") and the unsupported length is 5.818 meters (229.055"). The required speed is 0.1 meters per second (3.927 in/sec), and an acceleration of ±2.5 m/s2 (98.4 in/s2) is needed. The duty cycle is 8 hours per day, 5 days per week, and 50 weeks per year with an average of 10 cycles per hour. The life requirement is 20 years for the ball screw and 5 years for components. An additional requirement is that a stepper motor be used due to a preference of the electrical engineering department.
Next, we select the linear bearings for the x-axis. The primary requirements of this application are high load capacity and high stiffness. The application has a relatively long travel length of 5.500 meters (18 feet); however, the availability of 6- meter-length screws eliminates the need for butt joining. Low maintenance is an important requirement of this application. The result was the selection of 500 series ball profile rail linear guides.
With this selection made, the load on the ball screws can be calculated. Based on this loading, the NEFF KGF-D ball nut is selected as the starting point. This ball nut has an integral flange, integral wiper, and a DIN 69051 mounting. The ball screw has an accuracy of ±50 μm/300 mm accuracy.
Next, the life expectancy requirement is checked. Life is typically rated at L10, which represents the time after which 90% of ball screws will still perform. In this application, life expectancy is 1,035,752.6 years. The reason life is so high is that we selected the ball screw based on critical speed rather than life.
Testing the Proposed Design
Once you have selected your design based on the calculations, you need to test to make sure your premises are correct. The testing is designed to validate that what was proposed was actually delivered and, if that was not the case, to guide any corrective actions that may be required. Validation testing should be designed to answer questions such as:
• Does the finished product meet the design specification?
• Is the performance consistent, within experimental limits, with the theoretical calculations, and if not, by how much does it vary and why?
• Does the product provide the required level of reliability?
• What are the potential modes and points of failure for the product?
• How does the current solution compare to alternatives?
For large systems and machines, you may wish to begin with component testing before moving on to bench testing of subassemblies, and then finally to testing of the complete assembly. At each phase of testing, the test results should be reviewed and compared to the theoretical calculations to make sure that the design is on the right track, or consider reasonable opportunities for improvement. Testing is intended to reveal what we might have missed in our calculations and modeling.