Detecting a failed accumulator?


#1

My company, like many, makes a range of customized standard products for our customer base. I have implemented an RMC75 into this system to control the machine’s motion during operation. Recently, we built and shipped a machine from the US to Singapore and were plagued with performance problems. Our technical support team visited the customer and validated all electrical and mechanical systems and still couldn’t find the problem.

I was able to obtain plots of the machine running and I noticed that the control loops didn’t seem to be performing as well as they typically do on this machine. The system’s response was “off” especially when a disturbance was encountered. While participating on the “what to do next” conference call, I realized that the accumulator could possibly be the problem. After the team dug into the situation a bit further, we found indeed that the accumulator had not been charged. We chose to have our local hydraulics distributor charge the accumulator prior to shipping the units to us and unfortunately, they accidentally didn’t charge the unit.

Rather than debate the decision to have a vendor charge accumulators, I’m wondering if anyone has a best practice to test for a failed accumulator with a user program.


#2

I think the answer to this is going to be very application dependent. Given the freedom to spend whatever you want, you could put a pressure transducer on the nitrogen side of the accumulator and watch that. Almost as good would be a pressure switch on the nitrogen set a little below the pre-charge pressure. Look at the switch when you know that the hydraulic pressure has been vented to tank. If there is not enough pressure to make the switch, then the pre-charge is low.

You could write routines to observe the system response under some known conditions and possibly point to the accumulator like you did by looking at the plots. This is better than nothing and could generate a message to check the accumulator pre-charge.

Be aware that Nitrogen follows the perfect gas law so the steady state pressure on the nitrogen side will change with temperature so you can’t set your limits too close to the edge.