In this article, I want to talk about a simple way of determining the safest direction of an external radial load on the shaft of an axial-piston swashplate-type hydraulic pump.
Since the rotary group of a typical swashplate-type axial piston pump creates an asymmetrical drag (due to the fact that it is only half of the pistons that is inducing most of the braking torque) there is a significant amount of radial load on the shaft bearings, especially at high pressure and full displacement. The bearings are always properly sized to withstand these loads, but in certain applications, the driving mechanism can create an additional radial load on the shaft bearing - for example in the case of a belt drive or a direct gear drive, and in such situations, the direction of the external radial load on the pump input shaft is very important, because it can either add to or subtract from the internal radial load. Obviously, adding to the radial load will do nothing good to the shaft bearing's life expectancy, so a system designer must be very careful when designing a belt (or a gear) driven pump system, and it seems that simply consulting the respective data sheet may be not enough because there's often no information on this subject (aside from a generic "please contact the brand's representative for advice bla-bla-bla"). I'll say even more - if you dry "digging" a bit - you'll find that the information in catalogs can be downright contradictory!
Let me illustrate with a concrete example - let us take a Danfoss series 45 pump and a Kawasaki K3VL. Both of these are variable displacement axial piston pumps that have a similar construction in the sense that they have the same direction of the swash-plate angle. This means that the internal radial loads on the shaft bearing should be pointing in the same direction, and yet if we consult the respective datasheets for the recommended correct "external radial shaft load direction" we find the following:
You can see that the Kawasaki's catalog has the arrow pointing down and the Danfoss has the arrow pointing up! Very unhelpful, to say the least, don't you think?
So let's try and explore this subject and see if we can come up with some truth. When I need to visualize what is going on with the rotary-group-induced radial forces inside an axial piston pump I like imagining a pump with a single piston. I picture something like that in my mind (the pressure side of the imaginary valve plate is on the left):
First I imagine that I am freely rotating a single piston around the shaft axis with the help of a lever, welded to the shaft. Then I imagine that when the piston enters the pressure side of the valve plate it "starts to brake" - and this means that my lever is now pushing the piston in the direction it needs to go, and also pushing on the shaft bearing in the opposite direction (the green arrows). Then I keep rotating the lever, I visualize the last "braking position" of the piston - i.e. the place where it exits the "braking region" of the valve plate (the red arrows).
The angle between the green and the red arrows will be the "loaded sector" of the shaft bearing! It is that simple. More pistons mean more force, but the angle stays the same. If you do this "mental exercise" several times, you'll quickly see that the "internal radial load sector" kind of "lags" 90 degrees behind the pressure side of the valve plate. So, if you know which way the shaft is turning and where the pressure side of the valve plate is - you can immediately tell which way will the internal radial load push the shaft, and thus which way you should direct your belt drive to "fight" the internal radial load and optimize the bearing life.
It is that easy. Once again - just visualize what happens to a single piston - and you're golden! This also, sadly, means the Kawasaki catalog got it wrong...
Now, what about them closed-loop pumps then? Well, with closed-loop pumps things are a bit different because their valve plates can "brake" the pistons on both sides. And thus most of the pump manufacturers recommend directing external loads 90 degrees to the axis of the possible internal radial loads (which basically means aligning the external load parallel to the swashplate axis). The logic behind this is: "We'll be adding an external radial shaft load, yes, but at least we won't be aggravating any of the existing ones".
Obviously, if you know that a given closed-loop pump will be pumping in one direction only, you can override this generic recommendation and place your driving belt so that it "alleviates" the bearing.
See? The "single piston" trick makes it so easy! Now - don't use a belt drive (or a gear drive) unless you absolutely have to!