I have been writing a hydraulic servo simulator for the last few months. My motivation is education, like Dr Khalil, and the ability to do tech support a little faster. This is my third simulator. My first one was written in excel. It was big and clumsy and used very crude methods of solving integrating the differential equations. I wrote another in Mathcad, but every new release or version of Mathcad broke the simulator and not everyone has Mathcad so they can't run the simulator.
In mid July is start my third hydraulic simulator. This one is written in java so it should be able to run on any computer that supports java.

I just got the math to work last Friday. I calculate the
1. the control signal to the valve using a PID with feed forwards.
2. The valve spool
3. The pump flow as a function of the system pressure. I am assuming a time constant for the response of the pumps swash plate to the change in system pressure.
4. The volume of the gas in the accumulator. This is used to calculate the system pressure.
5. The pressure on the cap and rod side of the piston. This requires knowing the flow into or out of the each end of the cylinder. It also requires knowing the velocity and the position of the piston.
6. The resulting net force on the piston taking into account the piston.
7. The net force divided by the mass provides the acceleration. This in turn is integrated to provide the piston velocity and position.
Since solving this requires a system of differential equations I used Runge-Kutta at 0.0001 second intervals. The controller updates the target position and the control output every millisecond.

All together I have 15 data points to plot.
ftp://ftp.deltacompsys.com/public/Pictures/HydSimDev.png
I just got this going yesterday and I still have a lot of work to do. I don't have a means yet of plotting just the lines I am interested in but that will come. I know the graph is confusing with all the lines.

A brief description of the plot. The move is one that moves from position B ( 10 inches ) to position a ( 4 inches ). The diagonal lines that starts at 100 on the y scale that move to zero is the target and actual positions. The cyan line is the target position and the red line is the actual position.

There is another red line that stays close to the top. It is the system pressure. It drops from the set point of 1500 psi to about 1450 psi. If the plot ran longer that 1 second the system pressure would climb slow back toward 1500 psi.

The dark blue line that starts at about 86 on the y scale is the accumulator gas volume. The 86 means the accumulator is 86% gas. You can see the gas volume increases which causes the system pressure to decrease. The change in gas volume is simply the flow out to the valve minus the oil supplied by the pump. The initial volume is calculated using P0*V0=p1*v1. Where P0 is the precharge pressure and v0 is the size of the accumulator.

The green and pink line are the control signal from the controller PID and the valve position. The valve position lags a little bit because it has a 50 Hz response.

The orange line that starts at 65 and rises to about 87 precent of system pressure is the rod pressure. The cap pressure is about 45% of system pressure in the middle of the graph.

The blue and magenta lines that start at 50 and go do to 0 are the actual and target velocity. The target velocity is -10 inches per second. The actual velocity is not as great.

The red line on the bottom that increases to about 5% is the pump flow. The pump does not respond very quickly to the motion at all. That is because the proportional band of the pump is 300 psi which means that it will not come fully on stroke until the pressure drops from 1500 psi to 1200 psi. This highlights one of my big complaints about the garbage pumps you guys buy.

Finally there is a orange and yellow line that start at 50 and take dip then return to 50 and then rise and then return to 50. This is the target and actual acceleration. Actually the target acceleration starts and ends at 0 in/sec[sup]2[/sup] and dips -187.5 in/sec[sup]2[/sup] and then rise to +187.5 in/sec[sup]2[/sup]

This is my never ending project. It will change of time. I don't like all the dialog boxes. I can move the black vertical hairline but I don't display the values at the point. I need to do so. Eventually I would like to see the simulator run real time so the data scrolls across the screen. I also need to take into account piping and filters. At this time there are no loses for piping. I intend to do calculation similar to what stroup did to calculate the pipe loses but I also want to take into account the piping capacitance, especially for hose.

I know that most will not be able to understand but perhaps Dr. Medhat Khalil should be able to appreciate this since this is one of his interests.

Did I mention that I had to learn java?

Greeting, I can understand of course the effort required to solve this problem. What is the ultimate goal?

Peter Quipped:

quote:

The pump does not respond very quickly to the motion at all. That is because the proportional band of the pump is 300 psi which means that it will not come fully on stroke until the pressure drops from 1500 psi to 1200 psi. This highlights one of my big complaints about the garbage pumps you guys buy.

That's what you get when there is only a few in the world who no enough to give the type of feedback required to make pump manufacturers know there is a problem with their product.

If TRAINED/DEDICATED FLUID POWER PERSONS was the normal way to implement Fluid Power applications I believe the situation would change. There might be some Foot Dragger compamies that would go Belly Up but there would be others that would take the challenge and upgrade their equipment.

I found out years back that one person complianing about a need for changes to Fluid Power equipment got no respect and usually little or no response to questions about why a feature was not available to increase a Fluid Power products usefulness.

I guess you can see you have not convinced me that there is not a need for TRAINED/DEDICATED FLUID POWER PERSONS, Peter. As long as you keep helping me with fuel for the fire I will continue to stoke it. Thanks.

On a positive note, I applaud you for taking the time to work out the details for a Computer Program that will simplify the design of Motion Control circuits and eliminate a lot of rework those type circuits often require.

BTW, I hope you don't expect your effort will be MONETARILY rewarding. If the sales of Fluid Power training or software programs I've heard about are any indication you will have trouble buying Morning Coffee for over a month with the procedes without adding your own money.

Is that PESSIMISTIC or maybe just REALISTIC?

quote:

What is the ultimate goal?

Application support, education and marketing. Delta Computer Systems makes the RMC75 hydraulic controllers you have in your labs.

Two weeks ago we simulated a four corner press for compressing and baling hay to half its normal size for easier transportation. The code we generated worked right off with only one parameter needing to be adjusted. Simulation saves time and money. To do this we used the simpler linear second order simulator that is built into the motion controllers you have in your lab.

I was also in dire need of a skills update so I used this program to learn java. That is why the program looks crude and there is no fluff. The fluff will be added as I learn how to do so.

I often give presentations at a trade shows. I take these opportunities to show the flaws in manufacturers products and thinking. Now it will be easier show the flaws. I am not a politically correct person. I call them as I see them and anything or any company that makes my/our life more difficult is fair game. I try to be nice. I try to take a positive approach by raising the high bar but often that doesn't work. Speaking at trade shows is effective.
In the past I have criticized hydraulic valve manufacturers for selling non-linear valves as servo valves. Apparently enough customers got the message and applied enough pressure that one major valve manufacturer was convinced to make linear versions of their spools. Linear spools may cost a couple hundred dollars more but there is a pay back because the auto tuning makes it possible to tune the system better and keep it in tune.

So this simulator is also a hammer with which I will beat up the hydraulic suppliers. I will ask why I don't have their information in my data base. ( first I need to write the data base code

[quote=Budt]
BTW, I hope you don't expect your effort will be MONETARILY rewarding. If the sales of Fluid Power training or software programs I've heard about are any indication you will have trouble buying Morning Coffee for over a month with the procedes without adding your own money.
[/quote]
I have no illusions and I don't expect to make any money directly from this program. I don't even expect to sell it. Microsoft doesn't make any money off its browsers but Microsoft uses the browser to steer people in their direction. We will do the same.

This program is almost useless in a practical sense. Until I can get real models or equations from the manufacturers the results are just an estimate of what will happen and not suitable for real design although it may be better than the nothing everyone currently has. Currently the manufacturers do not provide differential equations or transfer function that describe how their parts work. Until that is done the hydraulic industry will be in the dark ages and no matter how skilled the engineer is he will still be making educated guesses. Believe me, I make that point at the trade shows.

quote:

I guess you can see you have not convinced me that there is not a need for TRAINED/DEDICATED FLUID POWER PERSONS.

So what should a person learn? I consider much of what is your books obsolete and counter balance valves are evil. Still there is much to know. Like picking the right devices and how to assemble them. This means welding and knowing fittings. One must pick the right oil and keep it clean. These are things I know little about but I acknowledge are necessary. Somehow these factors are rarely the cause of the problems I see.

I see people picking the cylinder diameters that are too small because they use V=Q/A, the infamous 'flow makes it go' equation. I see people using counter balance valves instead of blocking valves. Counter balance valves with regen is insane. I see people trying to accelerate huge masses very quickly and doing it and then complaining to me that their floor shakes. I see people choosing non-linear valves and wonder why the system doesn't tune up nicely. Even worse I see users trying to control pressure with overlapped spools. That happened just this last week and it was a distributor that specified the valve. The distributor told them the dead band compensator in the electronics would compensate for the dead band and it doesn't because the frequency response of an over lapped spool is 0 while it is in the dead band.

Budt, training people to get oil from point at to point B and keep it off the ground is one thing. The lack of engineering, math and physics skills is another and the lack of engineering data and tools is a third. The real problem is much bigger than keeping oil off the ground and knowing bang-bang or erector set hydraulics.

Bud, every year we have our three day distributor training at Delta. We cover these topics that are important to us because we know they don't get covered elsewhere. We have 25-35 people show up for these. We have two sets of training running in parallel. One is for the new people and is more of an introduction and training on the motion controllers. The advanced session covers more advanced applications, features and commands. I also give my presentation about the servo hydraulics. I think we have four people giving 15 minute presentations about their advanced applications too. During lunch and dinner there is a good exchange between the distributors of ideas how they did different projects and products they used.

good luck in your goals. You may review me may be in 4 monthes from now. I'm working on smoething similar. The first release will be about 4 months from now.
Thanks

Qupte Peter:

quote:

Currently the manufacturers do not provide differential equations or transfer function that describe how their parts work. Until that is done the hydraulic industry will be in the dark ages and no matter how skilled the engineer is he will still be making educated guesses.

Is it possible they don't really know the answer to your iquiry????? I still say the Industry is in the Dark Ages and until there is real feedback from MANY TRAINED/DEDICATED FLUID POWER PERSONS it will stay that way. Don't expect the manufacturers to jump on your request. I doubt you will hear anything until enough TRAINED/DEDICATED to FLUID POWER PERSONS speak so loud that they cannot be ignored. That, I believe, is why the Mechanical and Electrical fields have advanced so quickly over the years.
"So what should a person learn? I consider much of what is your books obsolete and counter balance valves are evil. Still there is much to know."
In the world of Motion Control you are correct in saying this, However Motion Control is a VERY, VERY, VERY small part of the Fluid Power field. I may live in a backward part of the world but I see less than 5% of Fluid Power installations using Motion Control. I want to see the other 95% have real TRAINING so the everyday Mundane Circuits don't Shake, Rattle and Roll, Spew fluid from way too many places and require a Heat Exchanger to keep them below 200 Deg.F.

On Counter Balance Valves, It would be more logical to voice your complaint to Bosch Rexroth, Eaton Vickers and Parker to name a few. These valvesmhave been around for years
and are on almost any Hydraulic circuit that has an Over Running Load especially if the actuator has verying speed. They perform very well when correctly applied but one must know how to apply them and that leaves out the UnTrained or the person with a one or two week school they attended 3 years back.

BTW, I have used Flow Regeneration to speed up several Cylinder circuits that had Counter Balance valves to resist running way actuators. Never had a problem after the old College of Hard Knocks Trial and Error period took effect and I learned how to correctly apply them.

"I see people trying to accelerate huge masses very quickly and doing it and then complaining to me that their floor shakes. I see people choosing non-linear valves and wonder why the system doesn't tune up nicely. Even worse I see users trying to control pressure with overlapped spools. That happened just this last week and it was a distributor that specified the valve."
Distributors have the same training as the rest of the Fluid Power world, they just get more of the old College of Hard Knocks quicker than persons in industry, and I can say that is from my own experience.
What you wrote is, SADLY, true and will remain that way until there are TRAINED/DEDICATED FLUID POWER PERSONS applying the equipment in a planned manner.
I agree the training is not about keeping oil off the ground, though that is a start in the right direction. It would eliminate the Drip Pans, So Called, on a lot of machines powered by hydraulics and save a lot of wasted oil and the cost's it brings about. The oil is not cheap but the clean up and disposal is even more expensive. Try to find a leak on a system that was installed by Tube-mac employees if you can. That is not to say their system is fool proof but when they install a system it does not leak.
"Bud, every year we have our three day distributor training at Delta."
WOW, a WHOLE THREE DAYS EVERY YEAR!!!!!!!!! I bet these trainees are teaching you stuff after that!!!! Sounds like Motion Control is as simple as the every day mundane Fluid Power stuff.
You still have not convinced me that Fluid Power has no need for TRAINED/DEDICATED persons that Design and Maintain Fluid Power installations in the same manner it is done in the Mechanical and Electrical world. In fact some of your statements sound like there is more of a need for these persons than I originally thought.

Peter - we met last year when I was helping a Yakima, WA company find a good controller.

I am wondering why you do not start with one of the many simulators that are out there already? Several have very flexible (and even IEEE standard-compliant) modeling languages and interfaces to generate custom solutions like you propose. I have seen so many attempts to make new simulators that get bogged down in the algorithms used, the user interface, pre and post-processing, etc., that little attention can be focused on the actual models. The simulator is pretty generic while the key is in creating good models of the actual components and effects.

Just a thought....

quote:

Originally posted by Dale Witt:
Peter - we met last year when I was helping a Yakima, WA company find a good controller.

Yes, I remember.

quote:

I am wondering why you do not start with one of the many simulators that are out there already?

Actually, we have 20Sim. This is my third simulator. My second one was written using Mathcad which never got as far as what I have got already. I am also limited it what I can do with a commercial simulator both in the features and distribution. I may turn this into a webstart or applet.

quote:

Several have very flexible (and even IEEE standard-compliant) modeling languages and interfaces to generate custom solutions like you propose.

What makes a modeling language IEEE standard compliant? If I write a simulator in java or C# then I have control.

quote:

I have seen so many attempts to make new simulators that get bogged down in the algorithms used, the user interface, pre and post-processing, etc., that little attention can be focused on the actual models.

Yes, I know. The algorithms and math are easy. Getting the model information from the manufacturers is the biggest problem and that will be the same no matter who's simulator I use. Personally the biggest challenge for me is writing the HMI. I don't do windows and I have been graphically challenged so this is also a learning project.

quote:

The simulator is pretty generic while the key is in creating good models of the actual components and effects.

Yes, and that problem is the same no matter which way I go.

quote:

Originally posted by Bud T:
Qupte Peter:

quote:

Currently the manufacturers do not provide differential equations or transfer function that describe how their parts work. Until that is done the hydraulic industry will be in the dark ages and no matter how skilled the engineer is he will still be making educated guesses.

Is it possible they don't really know the answer to your iquiry?????

The true engineers know what I am asking for. The business decision will be to ignore my request because I am the only asking for this data.....for now.

quote:

I still say the Industry is in the Dark Ages and until there is real feedback from MANY TRAINED/DEDICATED FLUID POWER PERSONS it will stay that way.

Stop whining. Join the 21 century and realize that force makes things move and pressure is just one part of the potential energy of oil.

quote:

Don't expect the manufacturers to jump on your request. I doubt you will hear anything until enough TRAINED/DEDICATED to FLUID POWER PERSONS speak so loud that they cannot be ignored. That, I believe, is why the Mechanical and Electrical fields have advanced so quickly over the years.

The hydraulic manufacturers will just have to realize that providing design information will make their product more likely to be used by designers. Except for a little documentation and testing it costs little.


"So what should a person learn? I consider much of what is your books obsolete and counter balance valves are evil. Still there is much to know."

quote:

In the world of Motion Control you are correct in saying this, However Motion Control is a VERY, VERY, VERY small part of the Fluid Power field. I may live in a backward part of the world but I see less than 5% of Fluid Power installations using Motion Control.

I suppose the bang-bang hydraulic systems are being replaced with bang-bang electrics.
Can't you see. I am in the no mans land of the battle between servo motors and hydraulics. If there wasn't servo hydraulics many more applications would go to the electric servos.

quote:

I want to see the other 95% have real TRAINING so the everyday Mundane Circuits don't Shake, Rattle and Roll, Spew fluid from way too many places and require a Heat Exchanger to keep them below 200 Deg.F.

It will be hard to compete with the smooth motion of a servo. It will be hard to compete with the flexibility a servo provides. You would send those 95% with real bang-bang training to compete against the servo guys? Good luck.

quote:

On Counter Balance Valves, It would be more logical to voice your complaint to Bosch Rexroth, Eaton Vickers and Parker to name a few. These valvesmhave been around for years
and are on almost any Hydraulic circuit that has an Over Running Load especially if the actuator has verying speed. They perform very well when correctly applied but one must know how to apply them and that leaves out the UnTrained or the person with a one or two week school they attended 3 years back.

Counter balance valves may work in a bang bang system. In a servo system there can be only one. That is only there servo valve can control flow. A counter balance valve interferes with the servo control.

quote:

BTW, I have used Flow Regeneration to speed up several Cylinder circuits that had Counter Balance valves to resist running way actuators. Never had a problem after the old College of Hard Knocks Trial and Error period took effect and I learned how to correctly apply them.

I am sure you have but not in a servo system. There is no control of the flow on the rod side of the cylinder. It is hard to stop quickly after extending.

Peter wrote:

quote:

It will be hard to compete with the smooth motion of a servo. It will be hard to compete with the flexibility a servo provides. You would send those 95% with real bang-bang training to compete against the servo guys? Good luck.

I'm not advocating any COMPETITION between Bang-Bang and Motion Control circuits. I see a Clear and Distinct Line between the two types of circuits. You know from experience I've called on you and your expertise in Motion Control and will do so anytime I am involved with a circuit that needs precise control. However, there are numerous circuits that no one in their right mind would consider using a Proportional or Servo Valve on unless they have extra money to burn and have the expertise to correctly install and service these circuits.

That is not to say the Proportional or Servo circuit would not perform equal to or better than a correctly designed Bang-Bang circuit. To me it is equal to making every Postal Delivery vehicle a Mercedes or Lexus. The Jeep type 4 Cylinder delivery trucks presently in use do an adequate job at far less expense up front and also for operating cost. Overkill can work well and some pay the extra expense but most opt for adequate and spend the extra money elsewhere.

At least this discussion is a look at the way Fluid Power is and has been implemented over the years. I for one think it stinks but I have been wrong before and another time will not send me away Whining.

I probably should take the lack of FEEDBACK on the subject as an indication that Peter is correct and that all Fluid Power Circuits should use Proportional or Servo Valves. Sure would cut the amount of inventory on most manufacturers shelves.

But, Then, The Dark Ages would be a LOT DARKER due to the lack of TRAINED/DEDICATED persons who not only don't know how to design and maintain Bang-Bang circuits but now are expected to understand Proportional and Servo circuits operation and maintenence.

Oh well, that's Life!

You should have seen some the cool stuff. One of the distributor showed a movie of a 6 degrees of freedom motion platform. I was told that is is the largest 6 degrees of motion platform in the world. It was amazing how smooth it moved even though the load changed.

There was also a rock crusher that might choke on a Hummer but could crush the average car like a college student crushes beer cans. The trick is that the crushing roll must be able to back off when a big rock or piece of solid metal goes through. The controller keeps the rolls parallel and from binding.

Another application is a 4 corner press the compresses hay. It keeps the 4 corners synchronized within about 0.004 inches until the load is hit. Then is stays synchronized within about .035 thousands. There is no accumulator to keep a steady system pressure.

Bud, you focus too much on the negative.

quote:

Bud, you focus too much on the negative

I don't call it negative Peter, just realistic. However from the copious amount of feedback generated it must all be in my head so the vendetta for TRAINED/DEDICATED FLUID POWER PERSONS is history, at least for me.

We had four distributors give 'show and tells' about their success stories. One distributor gave a presentation about steam turbine flow control. That inspired another distributor to get on the phone during a break. He got a PO for doing a steam turbine flow control. I don't know how he did but he did it but he did. The point is that if you get a lot of the best together to share ideas and techniques it can benefit all.

I 'blasted' the concept of 'flow makes it go' and 'pressure is resistance to flow'. I went through the steps that it takes to do one iteration of the simulator. Most could follow the steps when described on at a time but the solving all the equations in parallel was too much except for about 3 or 4 people. I showed how force makes the actuator go. How flow just equalizes pressure and pressure changes incrementally. There is no equation for pressure itself.

One of the more experienced hydraulic people also convinced me that counter balance valves are not 'evil' unless they are piloted off the cap side. Counter balance valves will work if internally piloted in a similar way as you showed in the synchronizing cylinders thread. However, they still cause a pressure drop proportional to the flow squared and make the system more non-linear than it already is.

I need to include the ability to simulate a system with a counter balance valve both internally and externally so I can show the difference.

quote:

I need to include the ability to simulate a system with a counter balance valve both internally and externally so I can show the difference.

Don't forget you can also get Counter Balance valves tat have both Internal and External Pilots in the same valve.

Internal makes for a smooth action while in motion while external removes all the back pressure when full force is required like in a Clamp circuit.