Caterpillar C32 Torque limiting

[Mike AA]

I am trying to find a solution to a problem we have been fighting for a while. We have 4 Caterpillar C32 engines connected to a 4 input 1 output gearbox. While running the engines all load up randomly. Some  times if we are trenching in hard ground or just hard materials we can get all the engines to load up to 100%. If we are in solfter easier to trench ground we get the random loading. I have a program running that has 1 primary throttle knob and 4 individual throttle/tuning knobs. The main throttle will go idle to full throttle and the individual throttles start at 100%. While runing the operator will turn down the throttle on any engine that is showing loaded more than others or a higher EGT. This somewhat works but we want to automate the tuning, I tried a few different programs and had another guy we hired who had some Guide experience and we struggled and didn't get that to work, I think it was most likely related to not being able to tune the PID well engough. We went back to the individual knobs.

I am now messing with limiting the torque of the engines to prevent them from using max torque and damaging the gearbox. Here is where I really am looking for the help, using TSC1 on the canbus for controlling the engine speed, I am sending 700-2000 rpm via SPN898 and can throttle up the engines no issue. I then started messing with the Torque limit, SPN518 recently to see if I could limit the torque that way. I installed the program on 4 engines and loaded up then dialed the torque down to 50% and saw  loads still higher than 50% but I don't really know if the LOAD, EEC2_x1, SPN 92, is original engine percent load or the reduced torque engine load. I changed EngOvrRdCtrlMode - SPN 695, I normally run set as 1 but changed to 2 to see if I could change the torque but lost all throttle control.

Has anyone done any engine torque limiting on caterpillar engines? I know there will be questions please ask away as I need to figure this out.

[samuelhj]

My first question would be, what are you using for feedback from the CAT?

As if I understand correct, you are interfacing the CAT's ECU via CAN bus which is reasonable, but keep in mind that the engine's ECU will have some ways to make sure the engine doesn't stall under load.

If you're running ALL four engines through a 4 to 1 gearbox, it means ALL engines will need some kind of feedback for adjusting what the output is. This is extremely time sensitive task you're doing. Especially not knowing what kind of load balancing gearbox you're using, this is going to be a whole new level of a headache.

However, you say that it works fine with tuning each engine for that precise load. While probably not good for neither the engines nor the gearbox this probably works for some time, but the whole system is still going to be under stress I'd assume. With a bit of a disclaimer I haven't seen the schematics for this and have no idea what type of gearbox you are using for this application, but marine applications use double input gears quite a lot so I'm sure there is a good solution to that part of the problem, which you've already found, at least the gearbox.

What Danfoss controller are you using for this application? And what does the plus1+ guide part for the engine control look like, roughly?

And last question, are you replacing an older controller with a new one? Or inventing this solution on your own? That's kiiiiind of important.

If you want, you can DM me some pictures of the setup and I can take a glance at it. This sounds like an interesting problem to solve.

[Lukey]

How does the gearbox work? does it require the same RPM from all engines or can they vary independently? the way i read your description it seems like they can vary.

Curious how did your initial PID work? 1PID or 4 (each engine)?
What's the setpoint, feedback and what is the output you were using.

From the information you've given my starting point would be to try having 4 independent PIDs, 1 for each motor.
The setpoint is your desired load/rpm/torque, feedback is that from the individual motor and the output is the throttle to that motor.

[pinias]

Hi Mike,

what is your background? I mean are you just electronic guy, mechatronic guy or just mechanic guy.

I am mechanical engineer with a master in electronics, if I would have to do something like that I will be monitoring the load percent of each engine.

that will be my 4 inputs, without any fancy control I just will find out the average load then just work on each engine to get that load, imagine a 3Phase incoming power the phases need to be balanced.

the picture attached shows the torque limit curve for the C7 to C32 CAT engines...from that you can said that the RPM´s will be between 1300 to 1800 RPM to be in the valid zone for torque limit, which makes sense.

it looks like just getting the average load then adjust RPM´s to get the loads evenly taking care of maximum or minimum RPM allowed is all what you need to do.

I hope this help

 

[Mike AA]

I will try to answer some questions and give more information that we have learned in the last few weeks.

Currently we decided to use math and my current control of individually controling the throttles and watch the loads and adjust by a step, in a stime period with a window, all values we can change on the display to tune the loading. On 2 machines now this is working really well.

To Clarify, the 4 CAT C32 are each hooked to a 5 speed transmission which then outputs to a gearbox that combines their inputs to 1 output. The gearbox is huge and requires two pieces of equipment or a very large excavator to unload and attach to the machine.

I am using a bunch of Danfoss controllers and IO to do this just due to how and where we are controlling everything. In the cab drive console I have a 24pin IO module reading the throttle knobs, in the electrical panel I have a MC090-10 which handles all the hydrostats, and a bunch of pressure transducers for the rest of the machine. On each pod is a MC050-118, a little over kill because we aren't recording any telemetry but we are using that on other stuff so it was simpler to keep using a computer we had on hand, plus it has enough outputs for the fuel, start and area light relays plus the 5 speed transmission coils. This MC050-118 takes my commands off the canbus and converts to what the CAT ECM needs for throttle and start and also relays engine info back to the cab displays.

We are inventing our own setup, nobody currently does what we do.

My initial PID ramped everything up and didn't adjust anything politely to say the lease. The new program we just threw on the second machine today is a very simplified version of a PID. We have the Position 1 load as the feedback and adjust 2,3,4 accordingly.

I am a jack of all trades, Electrician, Electronics, Computer programmer, Mechanic, Mechanical, Plumbing, Construction. When I work on our machines and anything with company or life or friends I help and can do virtually everything. Went to school for 1.5 years for electrical engineering, failed Chemistry and Calculus twice and lost my grant money. Got a full time job to make money and kept up my hobbies and experience doing electrical and programming to be where I am today.

Unfortunately the processes we run and materials we run every day, every foot of trench could be different than the next so there is not one load we always see. Today watching the feedback remotely I watched the load go from 35-45 for hours then up to 45-55 for a little while then back down to 30-35 until I lost connection. We have been able to keep the loads within 10% load of the master poisition 1 load.

Thanks for the feedback and questions and hopefully I answered your questions to make things make sense.

Mike