An Overheating, Closed-Loop Hydraulic System

By Quest Duperron, CFPIHM, Service Manager, Coastal Hydraulics

Closed-loop hydraulic systems are known for their reliability in industrial applications that require fast directional changes and long periods of service. However, even the most dependable systems can present challenging issues which can mislead and confuse seasoned technicians. This case study details one such instance where our team was stumped for days by an overheating issue with a closed-loop hydraulic system that utilizes these components. After days of testing, analyzing schematics, speculating, and consulting industry experts, we uncovered a hidden problem that highlighted the complexities of these systems.

Overview

The system in question utilizes a Denison Gold Cup Series piston pump. In this particular configuration, the pump included 2 additional internal pumps collectively driven by a 56 kW (75-horsepower) electric motor. The primary pump produces 214 lpm (56.5gallons per minute (gpm)). The first auxiliary pump provides 31.4 lpm (8.3gpm) for the servo and replenishing system, and the second provides 46.9lpm (12.4 gpm) for the charge system. This configuration powered a Hagglunds motor which was replaced a couple of years prior as a preventative measure after being in service for almost a decade. This motor operated for 2-3 shifts daily at times operating continuously. The Hagglunds motor is not a focal point in the discussion that follows, but is specifically mentioned due to its reputation for dependability and longevity throughout the years.

A Problem Emerges

Our troubleshooting journey began when we received a call from a longtime customer experiencing an overheating issue with one of its closed-loop systems. This operationally critical system powers a conveyor belt essential for daily functions. Despite the overheating issue, the customer emphasized that the system had otherwise been running perfectly. However, the issue needed to be resolved before the summer months as this would magnify the overheating issue. Furthermore, we wanted to ensure this was resolved early because it would lead to a shorter service life and potentially a premature catastrophic failure.

This was particularly significant to us because we had just recently installed this pump to replace the previous one a few weeks prior. The old pump had a damaged shaft seal which occurred when a coupler had been installed against the pump housing, destroying the seal’s race, and was sent out to be rebuilt. The fact that the pump was relatively new added a layer of urgency, as we wanted to prevent any permanent damage to the new pump.

Initial Inspection And Diagnosis

Upon arriving at the site, we began by verifying the temperature sensor’s readings using a handheld digital thermometer. We hoped a faulty sensor might be a simple solution; instead, this only confirmed the sensor’s accuracy. As we continued investigating, we found both the inlet and outlet of the heat exchanger were cool to the touch despite the elevated system temperature. This indicated to us that the heat exchanger was not performing as intended. As a result, we immediately began to focus our attention on the heat exchanger. Our first step was to isolate the heat exchanger and verify whether oil was being supplied to it. We disconnected the heat exchanger’s supply line, fed from the hot oil filter, and capped off the heat exchanger’s inlet. Then, we started the system and directed the hose over the reservoir to check for flow. We discovered that there was no oil flowing to the heat exchanger. This immediately raised our concern, prompting us to review the entire system’s plumbing against the provided schematic speculating that, on the off chance, we may have crossed a hose somewhere during installation. However, everything appeared to be routed correctly, so we moved on. Next, we turned our attention to the hot oil filter. Though unlikely, we thought the filter might be clogged causing the fluid to never enter the heat exchanger from the pump’s hot oil flushing valve (not illustrated). Once we disassembled the filter housing we discovered heavy contamination in the housing and a burst apart filter element.

Additionally, we found the filter housing’s internal 0.17 MPa (25 psi) bypass valve to be completely clogged, preventing any bypass flow. The bypass was designed for situations like this, where the element becomes too great of a restriction, obstructing flow, and allows the fluid to bypass around the element. This was an exciting discovery though tempered by the fact we did not have a replacement housing on hand which contained the internal bypass components. With that being the case, we decided to replace just the available filter element to restore flow and repeat our testing efforts.

Thankfully, the flow was restored to the heat exchanger’s supply line, so we reassembled the system to begin monitoring temperatures again at the heat exchanger. With the flow returned to the heat exchanger, we began to see the inlet’s temperature rise with a temperature drop across the heat exchanger. This indicated that the system was cooling again. We ultimately attributed the contamination to the previously installed pump’s time in service, determining that the issue was resolved and returned the system to its normal in-service operation.

The Overheating Returns

The following morning, we received an unexpected call: the overheating problem had returned. Naturally, we suspected that the filter housing was clogged again and expected to return only to replace the filter element again since the housing we needed was still on order. However, when we arrived and inspected the filter, we were surprised to find no new contamination in the filter element or housing. Confused by our findings, we repeated our earlier test by removing the hose from the hot oil filter to check for flow again. To our surprise, the flushing system was still pumping oil through the heat exchanger while the system’s temperature continued to rise.

It is important to note that we only observed the temperature climbing when the pump was stroked. The system’s temperatures would remain unchanged while the pump was in the neutral position. This nuance of a clue would eventually lead us to focus on the pump’s internal operation.

Delving Deeper

The next day, we meticulously inspected each component of the power unit and conducted various tests; after analyzing the system’s components, no obvious discrepancies were found. In our effort to be thorough, we proactively started testing some components that seemed to be unrelated to our issues, monitoring pressures and flow wherever possible ultimately unable to identify any discrepancies. Eventually, we reached out to a seasoned expert familiar with the Denison Gold Cup pumps. After sending schematics and a cumbersome explanation of our findings, we were advised to verify charge pressure and focus on the hot oil flushing valve’s operation.

With these recommendations, we verified the charge pressure 3.45 MPa (500 psi) and turned our focus to the hot oil flushing valve, which is designed to regulate oil flow to the heat exchanger from the off-service loop. After slowly backing off the valve’s relief, allowing more flow into the cooling circuit, we were excited to see the temperatures starting to drop. This, however,  was a short-lived victory as we soon had to abruptly stop the system when we heard the unmistakable sound of pump cavitation—a situation that could have led to serious permanent damage. Though the results were disappointing, it was this event that ultimately led us to a significant discovery: Case Pressure Replenishing, which should have prevented the pump from cavitating.

Understanding Case Pressure Replenishing

Case pressure replenishing, specific to these pumps, is achieved by allowing flow from the case drain circuit to enter the replenishing gallery and provide makeup flow to prevent cavitation. This is accomplished by the replenishing relief valve (8), a dual-area stepped poppet valve that allows back flow from the case to the replenishing gallery if the case pressure exceeds the combined spring and replenishing force on top of the replenishing relief valve. To take advantage of this feature, an additional arrangement is available, which is utilized in this configuration, for systems with long lines and/or large compressible areas to provide continuous cooled flow as an additional replenishing source.

Discovering The Culprit

With our newfound knowledge of Case Pressure Replenishing, the next logical questions were: How were we able to achieve cavitation? Why wasn’t the case pressure supplementing the replenishing gallery’s requirements? Normally, with these conditions and using a more standardized pump, we would start to consider returning the pump for repairs or sourcing a new replacement pump. However, some key factors made us reconsider this course of action:

The primary pump and both auxiliary pumps were individually flow tested the day before and passed. This suggested that each pump retained its integrity and was still capable of being functional.

The pump’s case drain should have been continuously supplied/flushed via the D1 port, which was supplied by the heat exchanger’s discharge, exiting the case through the D2 port, ultimately flowing back to the tank. This should have provided the required make-up flow preventing cavitation.

The hot oil flushing valve was also supplying the motor’s case flushing circuit which introduced the motor as a new factor broadening our troubleshooting focus.

These factors sent us back to the drawing board, looking for another cause and questioning the cavitation condition we produced with our previous adjustments of the hot oil flushing valve. We determined that the cavitation was likely caused by a 0.159 orifice at the case flush supply (D1) which was designed to only allow 4 gpm to be re-introduced to the casing. We speculated that during our test, the hot oil flushing valve was discharging more oil than we could replace through the orifice at D1. Note: Primarily, the fluid loss should have been supplemented by the auxiliary pumps (4,5) combined 78.4 lpm (20.7gpm) flow, which should have introduced plenty of replenishment fluid through the replenishing check valves (12, 12).

After deciding that the oil flushing valve was unlikely to be the culprit, we ventured into the tank where a 0.45 MPa (65 psi) check valve was tee’d into D1, the heat exchanger outlet, and the motor’s case flush supply. Our thought was that there might be issues with a worn or broken spring in the valve lowering the pressure rating, or that it might be stuck/lodged open. This would cause the replenishing circuit to drain entirely. The only issue was that the check valve was almost eight inches below the reservoir’s fluid level and not visible.

After pumping down the tank to gain sight of the valve’s discharge, noting that it was the correct valve and installed correctly, the system was started and stroked again. Immediately, we noticed that the check valve’s discharge, as well as the return filter’s discharge, was extremely heavy. This was a lucky break regarding the return filter’s discharge because it was located right next to the check valve and only visible with the lowered fluid level. After removing and inspecting the check valve, we confirmed that the valve was operating properly and unseated at 0.45 MPa (65 psi). The valve was then reinstalled, and we revisited the schematic with our new findings only to quickly realize that the most likely source of our heavy flow from both the return filter and check valve was likely supplied by the motor’s case flushing circuit. We determined that the Hagglunds motor was internally bypassing and supplying the case flushing circuit from the primary A and B circuits. While bypassing, the motor was not only generating excessive heat but also introducing the heated fluid into the pump’s case flushing circuit combating the hot oil flushing valve’s cooled fluid.

Conclusion

This case underscores the importance of thorough troubleshooting of closed-loop hydraulic systems, especially when dealing with intricate features including case pressure replenishing circuits. While the motor’s failure was hidden behind the functionality of the system, persistent investigation and consulting the right expertise led us to the correct diagnosis. With this conclusion, it became clear that the motor was the root cause of the heating issues, causing abnormalities in the pump. After replacing the motor, the system regained the ability to operate without the heating issues and continues to be in service to date.