OK, so last week we played with a Danfoss OSPC ON steering valve, and we saw something strange. Let me recap real quick:
We all understand that a modern orbital steering valve is a hydraulic follower, based on a rotary directional valve and an orbital hydraulic motor, in which the return flow of the follower motor is directed into the steering cylinder, which gives us the ability to volumetrically control the output with great precision and allows us to use the motor as a steering-wheel-driven pump for emergency steering.
This was all nice and tidy till we looked at the official Danfoss diagram:
and imagined that in a situation when we have no feed from the pump, and the steering wheel is turned by the driver, the back-flow from the steering cylinder should not be able to turn the steering wheel back because of the check valve in the P line. Yet, when we simulated the feedback pressure from the steering cylinder on a test bench, we saw that the steering wheel turns back fast and strong!
This can only mean one thing - the diagram does not show the whole picture, and even though we have a decent understanding of what a hydraulic follower is, we don't see all of the nuances of this concrete implementation. So let's find out then, shall we?! And while we do that, let us see if we can come up with a better schematic that explains exactly what is going on. Time to open a steering valve!
A note before I say anything else: today I will only be considering the directional part of the steering valve (i.e., the rotary spool valve), and will omit the orbital motor altogether. Just to keep this post short. I will look into the orbital in great detail next week. It is very interesting, but for now, we will concentrate our attention on the rotary spool valve. Consider the orbital part of the steering to be a simple fixed displacement motor - oil goes in - it turns, you turn the shaft - it pumps. A simple fixed displacement volumetric device.
Back to the parts now. If you look inside the body, you will see that there are essentially five circular connection areas (or rings, if you will), four of which correspond to the ports (P, T, L, and R), and the fifth would be the ring of seven holes that feed the orbital section:
The first thing we need to see is how the oil flows to the tank in our open-center steering valve when the spool valve is centered. So, let us see how the body ports project on the spool valve:
With the spool valve centered, the L and R ports are blocked, and all the holes in the "orbital ring "are closed off as well. Just like the schematic shows. The oil from the P port flows through the 24 2-mm orifices into the center of the spool valve assembly and then passes into the T port through the slot that houses the centering leaf springs. The orifices may seem tiny, but since there are a lot of them, they actually amount to a pretty decent cross-section. In this picture you can see the light shining through the tiny holes when the spool is centered:
Now let us push the spool valve into a commanded position and see how the oil is directed. This picture shows the spool turned all the way to the right (clockwise):
No light is shining through the P to T by-pass, which means it is closed now, and also the light in the L port holes tells us that it is connected to the tank - which once again perfectly coincides with the Danfoss diagram. And, finally, we can tell that the remaining 6 holes of the P ring are connected to half of the 12 holes in the orbital ring section, while the other half of them gets connected to the R port (the "in" and "out" of our orbital motor).
OK, so far this seems to be matching exactly the catalog schematic. The P to T passage is closed off, one steering cylinder port is vented to the tank and the other is connected to the P via the orbital motor. So... Once again - we are seeing that if we were to pressurize the work port in that position of the spool - there would be absolutely no way for the oil to flow back from the orbital motor into the valve (other than the internal leakage, of course). So how the hell could we see in our tests the steering wheel "push back" so violently?
The mystery gets unveiled when we realize that in order for the orbital motor to be "pushed back" by the work port pressure, its return line must be connected to the tank, and if we carefully inspect the P to T shut-off arrangement, we will see that there's more to it than meets the eye:
As it turns out, the P to T bypass is not an on/off directional valve, but rather a proportional valve that is carefully designed to gradually throttle the P to T passage as the central spool is shifting. And the closure happens after the directional part of the spool is "activated" (for lack of a better word). In other words - the spool first connects the steering lines and the orbital motor, and only then gradually closes the P to T by-pass.
Look at this picture:
The spool is shifted to the right, and you can clearly see the slots inside the holes showing the orbital section connected and the L port vented to the tank, and there's still light shining through the P to T slot! As long as the spool is slightly "above" the end of travel, the bleed to the tank will allow for the steering pressure to push the orbital motor back because there's a passage to the tank venting its back pressure!
If I were to draw a diagram of such a steering valve, I would prefer something like this to the official version (omitting the anti-peak and anti-cavitation valves for simplicity):
The traditional bullet-points: