Several days ago I read an interesting article about hydraulic balance of mechanical seals (the original article is from here, I am also putting a copy of it in the Library) . Very interesting material, definitely worth looking into. As I was finishing reading the article, an idea of applying the same principles to piston assembly of an axial piston pump was already buzzing in my head. Many thanks to mr. Mc Nally for the idea!
Earlier I talked about hydraulic balance and its use to reduce load forces on pump components. This article is a follow up, focused on a piston assembly of a swashplate type axial piston pump. The pictures show a classic piston shoe design, although the hydraulic balance principle is equally applied to the "inverted" piston design, where the "ball" is part of the shoe (Linde, Komatsu..), the solution that allows a bigger swash-plate angle and therefore a more compact pump design.
Let us take a closer look at the forces acting on a piston inside a working pump. Obviously we have the closing force, resulting from system pressure acting on the piston's end and the opening force acting on the slipper's cavity. But besides these two obvious forces there's more. Although the clearance between the contact surfaces of the slipper and the swashplate is very small, there's still radial oil flow and, consequently, a pressure drop in the gap area. Due to the very small clearance the flow is most likely laminar, and due to the fact that, as the radius increases, the flow cross section also increases, the pressure drop is not linear. This means that the shoe balancing area is, in fact, a little bigger than the cavity area alone.
There are two basic slipper designs - the simple design and the complex (or labyrinth) design (just one example as there are many variations). The "high" areas inside the balancing cavity of a complex piston design also contribute to the hydrostatic lift as the film of oil under them is pressurized.
I would really like to know why so many different shoe designs exist. The simple, single cavity no labyrinth design is the cheapest and the fastest to produce, and its efficiency and reliability is time proven on both open and closed circuit pumps by many, so if I were to design a pump I would stick to it. But, apparently, there are still plenty factories out there, who invent a new revolutionary shoe cavity design each time a new pump series is launched. Which can also be explained by the fact that creation of pump/motor components to a great extend is (A) an empirical process, and (B) a process greatly influenced by company tradition, which is no shame at all. So if someone wants to go nuts machining away all sorts of cavities and cavity-connecting labyrinths in a piston shoe, I say let them! In fact, nothing can explain better this point than this abstract from (1):
"The general lack of basic principles for designing reliable components for hydraulic pumps and motors requires a strategy of applying experience and full scale testing. Hydraulic pump and motor components are very sparsely lubricated though they are immersed in fluid. They therefore slide in "boundary lubrication", a regime for which there are no methods for predicting, or even estimating product life or frictional performance. The testing of sub-components in bench tests or in accelerated tests produces new uncertainties since the role of the many variables that control wear and scuffing are not well known. The best design procedure involves full scale testing of components with all of the recirculating contaminants, vibrations and misalignments included..." (highlighted by me)
Now, to dress this theory up in numbers and to find out the balance ratio of modern piston designs (out of pure scientific curiosity, of course), I swept shelves in the warehouse and took measurements of some random pistons (various brands, both closed and open circuit), then made some calculations and combined the results into this nice little table. For a complex design piston the measurements were taken in the following manner. All measurements are in millimeters. To calculate the contact area lift, an estimated value of 30% of the total area of the outer ring surface was used. If you ask me where did I get the 30% figure I'll tell you it's a hunch, based on the McNally's article.
In the "balancing area" column you have the area of the shoe cavity directly exposed to the system pressure (the surface area of the "labyrinth walls" of the complex design shoe was not taken into consideration, as the fluid film under the "walls" is pressurized thus equally contributing to the lifting force) plus 30% of the total area of the outer ring. Balance ratio is the shoe balancing area divided by piston head area.
|Short Pump Reference
|Piston Head Area, sq.cm
|Balancing Area, sq. cm
As you can see, different brands and designs all wound up with the balance ratio of around 0.9, (the only exception being the A10VO028, with a smaller ratio of 0.8, which I am guessing can be explained by the small size).
You may also see that complex shoe design prevails. Another advantage for the design I might think of is, probably, the fact that the narrower outer ring of the complex design, when scored by contaminant particles, affects the balance ratio to a smaller extent in comparison to the simple design with its relatively wide ring surface. But hey, check out Sauer Danfoss series 90 closed circuit pumps. Extremely reliable and time proven units that have used single cavity design pistons for decades and seem to have done just fine.
It is also worth mentioning that aside direct acting forces caused by the system pressure (and the case pressure) there are other forces contributing to the piston shoe lift. There's the hydrodynamic lift, caused by the fluid trying to get under the shoe as it moves along the swashing plate. There are all sorts of vibrations and misalignments. There's the centrifugal force, friction forces between the cylinder block and the piston itself, and probably an assload of other forces I can't think of at the moment but pretty sure they exist. All of them being compensated for by the hydraulic balance.
I can only imagine how cool all the empirical work and failed attempts to design a reliable piston must have been! Pump designers, I salute you! Keep up the good work for us, hydraulics-junkies, to always have something to back-engineer...
1. Hydraulic Failure Analysis
fluid, components and system effects
George E. Totten, David. K. Wills, Dierk G. Feldmann.