A new mathematical algorithm from System Seals removes the guesswork in selecting cylinder guidance systems while also reducing seal failure.

By Kurt Sassmannshausen, System Seals Product Development Manager

Failing to properly calculate guide-band loads, rod deflections and the stress and strains distributed across cylinder guidance systems is a leading cause for guide-band fracturing and seal failure. Properly designing a cylinder guidance system is essential for maintaining metal clearances, and preventing overloads and equipment damage.

But accurately measuring guidance loads requires more than just guesswork. In most cases, the load is completely unknown and guide bands are often designed on assumptions. Many seal manufacturers mistakenly presume that stress is evenly distributed across the guide bushing, but a closer inspection reveals something else entirely.

Developing a new guidance model

In a fixed cylinder-end system, the amount of deflection in the guide bands increases as the side load increases. This correlation intensifies off-center driving and rod bending, which in turn, increases the load until an equilibrium can be achieved.

To better understand these relationships in a specific cylinder, and to optimize the guidance systems, System Seals in Cleveland has developed a new iterative algorithm that determines the side load, stress and strains distribution across an entire selection of guide bands. An iterative algorithm is a mathematical procedure that generates a sequence of improving solutions for a group of problems with several variables that impact each other.

The key output parameters of this algorithm are stress and strain over the guide bands and the total length of guide band in contact with the rod. These values are necessary to determine the optimum height needed for the guide band.

This is an ideal approach to understand the dynamic behaviors of guide-bands under load. The formula considers tolerances, cylinder weight and the couples created from bearing friction. To confirm the accuracy of the algorithm, Finite Element Analysis models were developed with the same variables and loads. Repeated modeling revealed that a height past which the guide band does not contact the rod provides no support whatsoever.

Understanding this concept is critical. Adding more guide band to accommodate high loads when the band is already at its optimum length is pointless. But without proper analysis, it is impossible to know if the band height is properly optimized.

Instead of creating a new solid model for each application, as necessary in Finite Element Analysis, the new algorithm pinpoints bearing stress, strain and size solutions based on solving multiple equations simultaneously with only basic dimensional inputs. Because these equations interact with one another, simple algebraic solutions don’t work. A numerical integrator, or iterative method, is required.

Testing conventional wisdom
In one test sample, the guidance of an entire system was located inside the seal groove. In this case two to three 25-mm bands are typically recommended. But the results showed that only 33 mm of the guidance was actually used, based on angular deflections of the system over the guide band. Although the basic calculation suggests that this design would be suitable, it clearly was not.

In a second test, a system was examined when a portion of the guidance was shifted outside of the seal groove. There was a 25-mm gap between the inboard and outboard guide bands to allow for a seal groove of 15 mm and 5-mm lands on each side.

The outboard material was a thermoplastic blend with a cross section equal to the main guide band. The test revealed a high load on the outboard guide band that created stresses higher than its yield. But a third test, where the cross section was reduced, the load was distributed to the point where the outboard guide band carried 16 percent of the bearing load, and yielding stopped in all of the guide bands. It was a clear example of a guidance system optimized by position, dimensions and choice of material.

Fixed-cylinder and free-rod end
In another test, systems with fixed cylinder ends and free-rod ends where examined. This included rams, presses and AGC-type cylinders. In this category, a standard 1,050-mm bore with a 960-mm rod AGC cylinder was used for the analysis.

The traditional approach is to install all of the guidance internal to the seals. Here, the analysis showed that about 70-mm of axial contact height occurred on the guide band. This means that only 75 mm is actually needed.

However, the leading edge of the guide band experienced stress values greater than its yield, so adding more guide bands would not have solved the problem. Only 69 mm of bearing height contacted the rod initially. After yielding set in, the bearing deformed to a larger contact area, and this approach would likely have resulted in a variety of failures that could have been avoided with customization.

An alternative design is to incorporate a guide band in the outboard position made of an engineered thermoplastic with an optimized cross section and height of 3.65 and 15 mm, respectively. This results in a balanced load distribution, allowing peak stresses in both bearings to remain below the yield values and the outboard guide band to carry about 15% of the load.

Conclusion
Traditional methods for estimating guide band stresses often fall short from misconceptions that loads distribute evenly. However, advanced methods such as Finite Element Analysis require resources that are not always available. Accurately predicting guide band loads is critical in preventing cylinder and seal failures. By using an iterative algorithm, the shortfalls of simplified assumptions can be avoided.

System Seals
www.systemseals.com