Insane Hydraulics

Site theme image

Negative Flow Control vs Load Sensing - Which One is Better?

Today I want to take a quick look at negative flow control (NFC) hydraulic systems, and talk about their main advantage over the classic closed center load sensing (LS) systems.

I know a lot of mechanics who understand how both the NFC and the LS work, but when I ask them why would one choose one over the other, or better put - what are the things that one system does better than the other one - I rarely get a reply. So, I decided to make a quick post to this very interesting (in my opinion) topic. Please note, that I greatly simplify things here.

Let's recall the principles.

Closed Center Load Sensing System

The most famous hydraulic circuit in the world. We all know and love it. It senses the delta P over the flow controlling element and tells it to the pump (or a compensator), who, in its turn, does its best to keep it stable no matter what pressure our work function should demand.

Negative Flow Control System

This is also a classic design, that has existed for a long time. It revolves around a variable displacement pump with the proportional displacement control with negative characteristic, and a specially designed directional control valve, which has an orifice downstream its central (by-pass) gallery creating the back-pressure to pilot the negative flow controller of the pump, and which also has spools designed to carefully throttle the connection between the P gallery and the by-pass gallery when they are shifted. When a spool shifts - the flow in the central gallery goes down and so does the orifice-induced pilot pressure - and the flow of the pump happily goes up. Simple and functional.

So, both systems work, and both systems apparently serve the same purpose - provide actuators with the flow when the DCV spools are moved - but if I were to name the one key difference between them, I would say that it is stiffness.

Yep - stiffness is the word of the day, perhaps for lack of a better term... Let me explain what I mean by that. When I say stiffness, I refer to the way a hydraulic system's flow reacts to a load increase. In a nutshell - a stiff system keeps the flow stable, and a more damped system lets the flow go down.

A classic closed center load sensing system is inherently very stiff because it is designed to keep function flow stable. This is perfect for a lot of things, but when you deal with inertial loads that are controlled by constantly changing inputs (read - human operator at controls), such systems are also inherently unstable.

Imagine an excavator boom that is rotating at a constant slow speed. An increase in the operator joystick input will, obviously, cause the LS system to respond and immediately "inject" more flow, however, the heavy boom will not be able to pick up the speed as fast as the hydraulic system flow increase, and so, naturally, due to the fact that the LS system will want to "push through" the desired liters per minute "no matter what" - most likely the flow increase will be accompanied with a steep pressure surge, which in its turn has the potential of growing into a very-hard-to-get-rid-of self-oscillation.

I bet if you used the ubiquitous PVG32 (a go-to closed center LS solution for everyone who works with Danfoss) in a crane-like system - you had to eventually replace the flow control spool in the rotation circuit with a pressure control spool, simply because instead of the expected "z-z-z" the boom rotation was doing "ziu-ziu-ziu"... Or maybe even went as far as installing an orifice between the rotation lines.

Anyhow - my point is - a stiff hydraulic system, that targets steady flow, is prone to self-oscillations when it is used in human operator-controlled systems (constantly adjusted control input) coupled with inertial loads.

Now, the NFC system, on the other hand, is inherently damped, because there's a controlled leakage path between the pressure and the by-pass galleries, which makes it perfect for damping any load increases. This is less efficient but offers a much more stable control, and for a system that is controlled by an operator, this is the most fundamental requirement.

Another thing, characteristic only for an NFC system would be the operator's "feel" of the machine. In an LS system things are simple - the joystick travel will always correspond to roughly the same actuator speed, but in an NFC system - the position of the joystick for the same speed will vary with the load. This means that subjectively, an experienced operator can assess the load with muscle memory - hence the better "feel". The classic LS just can't do that.

These are the basics, of course, and if you look at modern excavators or forest cranes, you will see that LS systems are as common as NFCs, which means that eventually engineers managed to "hack" an LS capable of stable performance, but still, at least in theory, the NFC offers better stability, better "feel", and a possibility of controlling force of the actuators, which, once again in theory, makes it a better choice for human operator controlled dynamic hydraulic systems.