Would like to hear opinions about the capabilites of a mechanical system

[Ben]

Hi All,

I am curious if the oscillation rate I see in a mechanical system could be a clue as to how fast I can run test cycles.

The plot shown above shows the drive signal (shown in red) being applied to a third-party hydraulic servo control board. That board controls servo valves that control hydraulic rams.

The black trace shows the stress feedback to the thrid-party servo board.

my customer is working the manuf. of the servo board to try and tune the PID better but I am curious if the frequency of those oscillations are a clue as to how fast I can run the actuator if I crank-up the frequency of my control signal.

My gut tells me "Ben it is obviously capable of moving that fast, just look at the frequency of the strees signal." But hydraulic are outside my personal area of expertise.

If I zoom in and estimate those "oscillations" in the black trace are just above 2 Hz. I would like to drive the test system at 2Hz max.

So can anyone comment if the frequency of those oscillations can be a clue as to how fast the actuators can run?

I appreciate your comments or insights.

Thank you!

Ben  

[Bob_Schor]

Disclaimer -- I am not an Engineer.  However, when I see "ringing", I think "resonance", and when I think of stimulating at the resonance frequency, I think of the Tacoma Narrows Bridge catastrophe ...

Bob Schor

[Ben]

---

@Bob_Schor wrote:

Disclaimer -- I am not an Engineer.  However, when I see "ringing", I think "resonance", and when I think of stimulating at the resonance frequency, I think of the Tacoma Narrows Bridge catastrophe ...

Bob Schor

---

Good point Bob!

At over a million pounds, that would be bad.

Associated thought...

Spoiler

Anyone can screw up but for a disaster, an engineer is usually invloved.

Ben

[BowenM]

I don't know your system... but one thing that is easy to forget in mechanical systems when you are talking driving frequency is that you need to overcome inertia.  This means that often the maximum frequency you can drive is, f = f (s, a) where s = system (servos, etc) and a = amplitude you are driving to.

[RavensFan]

I'd be curious to know what exactly is the cause of that 2 Hz oscillation.  Perhaps some shuttling of a pressure relief valve?  If it is caused by something within the hydraulic pumps that pressurize the system?  But I'd think that would show up at a higher frequency than 2 Hz.  If someone who understands the hydraulic system looked at this, could they find a way to put some sort of "snubber" in to smooth out the 2Hz oscillation.

If you do any experiments, I'd increase your rate very slowly.  See if the reaction on the 2Hz feedback gets worse.  My gut feel is if that 2 Hz oscillation persists, you're not going to be able to run your input anywhere near 2 Hz.

[BowenM]

---

@RavensFan wrote:

I'd be curious to know what exactly is the cause of that 2 Hz osicallation. 

---

Because I've seen it way too many times... my first guess would be a poorly tuned PID 

[Ben]

Thank you very much for opinions!

---

@BowenM wrote:

---

...

---

Because I've seen it way too many times... my first guess would be a poorly tuned PID 

---

A double thank you for that!

When I started this project (developed by one company, "fixed" (cough cough hack hack) by another and then I was brought in with the itro "It has never worked,... keeps breaking the DUT!) as I am standing in a puddle of hydraulic fluid. I suggested they fix all of the leaks which cleaned up the action enough that we could see what I presented above.

I was "shooting from the hip" when I aksed them to coantct the baord manuf. of the servo control board.

Your post above is wonderful to read and hopefully we are on the right track.

Spoiler

Not realted this project but since Bill mentioned gradually increasing the cycle rate very carefully.

I was asked to help get (yet another) application developed by someone else working. It tested the bumpers used under railcars (if you watch a train passing at a road crossing you may notice orange bumpers in the suspension) by crushing them with a hydraulic press. I wa watching the test run for the first time after I made changes and...

out of the corner of my eye, I could see the customer assuming a posture similar to a "crane stance" with his arms covering his head and raising his leg to shield his body. It worked, and I am still around.

And then again there was the time when I was delivering an application used for calibrating torch wrenches. When it was time to run the calibration in automatic mode for a wrench rated at 1000 ft-lbs, it was me standing behind the pillar!

Again thank you!

If other have an opinion they would like to share please do. And since I already took this thread off-topic, if you have crazy stories of final testing of dangerous applications, let em fly.

Ben

[Ben]

2.5 months later...

I am updating this thread just to offer a follow-up.

The customer did not want to touch the tuning of the servo card because there were too many factors involved.

The original supplier did not want to put an engineer on a plane so the problem lingered. I offered to try to tune the PID and lacking any other options, they took me up on the offer.

The tuning was not as complicated as the customer had thought because there were not seven factors involved.. rather it was set-up as a "PI" control loop. After shutting down the "I" term and slowly walking up the "P" term we had decent performance with only a slight over-shoot near the top of the sine wave. Turning on the "I" term and bringing it up slowly we achieved good results and the customer sent me away saying "I have to go tell the boss the good news."

We did some experimenting to make sure it worked OK over the entire range of frequencies and amplitudes. The customer was a bit more ambitious than I and he cranked up the numbers to see if the system would keep up with a 2Hz cycle.

At low amplitude, around 100K lbs. of force, the hydraulics were able to follow the command signal. When it attempted  a 1/2 million pounds of force a t 2Hz, the hydraulics were not able to keep up. Not being a hydraulics guy, I can only speculate the accumulator and or pumps are just not large enough.

Thank you for your opinions!

They helped!

Ben

[Ben]

Just short of a year after my previous post...

The customer wants to double the speed now that the system has been running without breaking things.

A hydraulic group has run the numbers and proposed upgrading almost all of the hydraulics. The one thing that presents a curve-ball to me is they want to use two servo valves instead of just one!

Would anyone care to offer any suggestion on how to go about tuning two PIDs each running one of the hydraulic servo valves?

I appreciate any suggestions or food for thought.

Ben

[BowenM]

So earlier today you posted a link to some Mainstream Preppers forum... which made me think of you... which in turn made me wonder if you'd ever gotten this solved.

Sadly - I missed your last two updates at the time.  Glad you fixed the original problem!  Did you ever get the two separate PIDs working together?